MadVR in use with LG OLED Thread [Archive] - Page 2

mark0077

14th November 2020, 03:02

I have tried a number of months without interpolation on my LG C9, after many years of using SVP, I have to say I don't know how people can be happy to watch 24p on a modern/good OLED without some compensation for motion/judderyness. With BFI mode enabled on the C9 at least, its too flickery as it only has an off/on mode. The "on" mode is very flickery compared to the low mode on the C10 and newer. (I welcome C10 inputs on this discussion when set to low).

We all know OLED is "perfectly" juddery as the OLED is doing as its supposed to do in 24p mode and we can get attached to that as being the 'directors intent'. I'm all for that, but I highly doubt that exactness / judderyness is what the director intended (how will we ever know eh?).

I really like SVP and its varying ranges of smoothness it allows, to over come this. Now it allows a pure avisynth_filter based way of engaging it, without the need for ffdshow. I think the whole "soap opera affect" is really over emphasised. This SVP smoothness looks fantastic to me and all of my relatives, SVP being engaged in a medium smoothness mode. It shouldn't be shunned so badly as "soap opera affect" and dismissed so badly/quickly.

In a way its like saying upscaling is "soap opera affect" (i.e. this is not natural, of course it isn't natural and that's the point, its to get as close to what the director intended, and that's not always exactly what's in the content/blu-ray, he/she probably didn't intend us to see a juddery image of 24p on an OLED right, just like he/she didn't intend us upscaling the resolution in crazy/great ways that we can do with modern hardware??)

I really look forward to this forum discussion move past it being "soap opera affect" and move into looking at ways to make our resulting image moving towards the true directors intent, which we all know is not a juddery 24p which a lot of us are seeing on these super modern OLEDs.

I'm not sure what the end solution will be though SVP and some tweaks to its settings (everyone is welcome to share their settings), or madVR (down the road possibly).

huhn

14th November 2020, 03:08

they clearly do even an OLED is slow compared to an CRT and they are not that "long" gone as mastering devices.

VBB

14th November 2020, 04:11

Next time you go to the movie theater (whenever that may be...), look for stutter in pans and fast action. It's there. That's the nature of 24fps, unfortunately, and since we're so used to it, anything else will look "unnatural".

SamuriHL

14th November 2020, 05:08

Correct. It's why people freaked out with the Hobbit movies when Jackson filmed them at 48fps. Which i was personally elated about and loved it. But it made a lot of people REALLY uneasy.

pirlouy

14th November 2020, 18:58

mark0077

14th November 2020, 21:04

@mark0077: Sorry I did not read all the thread. You're saying you use SVP on a LG Oled screen ?
Why don't you use the integrated motion interpolation from the TV ?

I also don't like the judder for 24fps content, but I use the one from my Samsung TV and I don't need SVP.
The advantage of using the one from Samsung, it uses its 120hz panel to convert the 24fps content, and you can adjust the fluidity to avoid soap opera effect.

I ask the question because it would be very disappointing if LG didn't have a motion interpolation good algorithm like the one from Samsung TV.

In PC mode I don't think the TV's interpolation can be used. It disables all of those features to keep latency low I guess. I would assume the TV's interpolation couldn't beat what a pc could do, similar to upscaling I guess. But I have not compared them. Something to try though, thanks.

Emulgator

14th November 2020, 21:08

To judge the unwanted side effects of (LG and other's) motion interpolation thoroughly I can suggest simple test patterns.
One can just generate blinking patterns of dots, or fences crossmoving. Very enlightening.
Or use a Spears & Munsil BD/UHD disc (Stacey Spears being a member here).

With the advent of AI crossmoving parts may come out more convincing.

In the end it seems to depend what one's vision algos are trained for.
Prefer smooth motion and tolerate such faults vs. finding them too distracting.

I find myself trying LG's motion interpolation (OLED55E6V) on and off,
just to find that some scenes might just benefit while others are definitely damaged.

A valid attempt for treatment of 24p content on large screens might be:
Introduce the correct amount of motion blur to make too short shutter angles consumable.

Pixar have demonstrated that perfectly:
Their short snowman-in-a-globe episode ("Knick Knack") had been rendered in 24p, has plenty of rapid motion, particles and stuff.
If you freeze a frame of the snowflake whirl you become aware how much love has gone into designing that motion blur.
Being from 1989 it still worked perfectly for me in 2019, BD 23.976p on a standard speed Loewe Connect 32" (no HFR interpolation).
Big thanks, John Lasseter et al.

This means there is no imperative to interpolate faulty frames when the vision algo is able to do that work on correct motion blurs.

quietvoid

17th November 2020, 20:27

To avoid cluttering the AVS thread more, what's the opinion here on the current saturation levels in high nit, HDR playback?
I've personally never noticed a difference for regular movies without desaturation, since the display is at DCI-P3.
Even demos, there aren't many scenes where desaturating makes a difference.

So all this testing for "correctness" seems very much focused at making low nit tonemapping desaturate much more than the HDR intent.
In my opinion, it feels like desaturating more for low nit projectors actually goes against the graded HDR master, which intends to display the color at the high brightness it should be.

Thoughts?

SamuriHL

17th November 2020, 20:57

While it's probably more noticeable for those with lower nit levels, we are not immune to the issue they're working on fixing. I don't pretend to understand all the details behind it all, but even when you have 700-800 nits, if you have very high nit content you have to compress that to fit below our real display nits. So even though it may or may not be as apparent, we're still dealing with the issue because we are in effect changing the brightness by definition. And indeed, I have played with the desat options on some higher nit content and can see very clearly even though I'm outputting 700 nits in HDR that saturation is impacted. What concerns me about that is even when doing passthrough to our display, we don't necessarily know how the tone mapping algorithm they're using handles this. madshi seems to think they're using dumb desat across the board in most cases. Ultimately whether you think you're impacted or not, the question is valid....since you don't have a 10k nit display, how do you KNOW you have it right? And from the content I've looked at so far, this seems more than a theoretical discussion....even for us.

VBB

17th November 2020, 21:29

I've been using HDR pass-through since I first got my OLED, but I've always wondered how much of a difference there really is. My 960 is not powerful enough, so I'll have to wait a while longer. I don't watch a lot of HDR content, but when I do, it looks so good that I don't feel like I'm missing anything. The only thing I did back when was to calibrate the high 2-point, and I disabled DTM and lowered color from 55 to 50 (this is now default on the newer models).

QBhd

17th November 2020, 23:05

I am fairly sure there are three different tonemapping curves in the C8. These are triggered by the MDL of the video being watched. The reason for my thoughts here are based on the 20-pt whitepoint settings... Looking at the second last code value we can trigger 3 different ones. 669,696, and 713. Most video seems to have an MDL of 1000 and there are some with 4000. The code value of a 1000 MDL video is 696 and the 4000 MDL video has a code value of 713... This is with HDR passthrough.

So how does this relate to madVR? Well it seems that when you use pixel shaders and output in HDR you are actually setting an MDL in the output with the "display peak luminance". I tracked down where the different code values switch.

DPL = display peak luminance
CV19 = the second last code value of the 20-pt settings

DPL 20-760 ==> CV19 is 669
DPL 760-2500 ==> CV19 is 696
DPL 2501-10000 ==> CV19 is 713

I keep seeing people saying that there is a way to disable the LG's tonemapping by setting the DPL to something like 700, but I am pretty sure every panel will have tonemapping at all times since the HDR into the TV needs to be mapped to the abilities of the panel.

Am I missing something here? I just think that capping the DPL to 700 is missing some of the potential for the top end highlights and for no reason since the TV will tonemap regardless.

QB

quietvoid

17th November 2020, 23:09

It is pretty complicated to desaturate even at 700 nits indeed, especially when we're talking about 2000+ nits source content.
I've compared the S&M UHD benchmark with madVR and the TV makes little effort to desaturate in order to retain more detail/contrast. I've posted this before but here are pertinent photos: https://slow.pics/c/fyivVAOB
The TV just displays a colored blob :)

I've been using HDR pass-through since I first got my OLED, but I've always wondered how much of a difference there really is. My 960 is not powerful enough, so I'll have to wait a while longer. I don't watch a lot of HDR content, but when I do, it looks so good that I don't feel like I'm missing anything. The only thing I did back when was to calibrate the high 2-point, and I disabled DTM and lowered color from 55 to 50 (this is now default on the newer models).
You're very much missing out, I watched Parasite (2019) straight from Kodi once and was disappointed when windows and skies were just plain clipped white :(
These days I'm too lazy to boot into Windows to use madVR though.. so Dolby Vision it is.

Am I missing something here? I just think that capping the DPL to 700 is missing some of the potential for the top end highlights and for no reason since the TV will tonemap regardless.

In the beginning it was established that 760 was the max nits one should set to have the highest brightness, however with panel variance etc. people just left it at 700 IIRC, to avoid surprises.

VBB

17th November 2020, 23:10

This is why it would be fantastic for future models to have an option (outside of the service menu, preferably) to disable tone mapping altogether, like what HGIG does for gaming.

quietvoid

17th November 2020, 23:32

This is why it would be fantastic for future models to have an option (outside of the service menu, preferably) to disable tone mapping altogether, like what HGIG does for gaming.

For what it's worth, the 2019 models (and above?) can set their own custom tonemap curves, though the color is probably still mapped by the TV.

SamuriHL

17th November 2020, 23:35

I keep seeing people saying that there is a way to disable the LG's tonemapping by setting the DPL to something like 700, but I am pretty sure every panel will have tonemapping at all times since the HDR into the TV needs to be mapped to the abilities of the panel.

Am I missing something here? I just think that capping the DPL to 700 is missing some of the potential for the top end highlights and for no reason since the TV will tonemap regardless.

QB

I don't believe anyone has tried to make the claim that using a ceiling of 700 nits DISABLES tone mapping. At least I'm not aware of anyone making that claim. What our thoughts are on this is that it will use the least aggressive tone curve. Which is about the best case we can get.

SamuriHL

17th November 2020, 23:38

This is why it would be fantastic for future models to have an option (outside of the service menu, preferably) to disable tone mapping altogether, like what HGIG does for gaming.

Seriously. Better than that, however, would be a 4000 nit panel. LOL

VBB

17th November 2020, 23:44

OLED will be long dead before it ever reaches such high nit levels, but you know that ;)

Makes me wonder what 2021 will bring, and if it will be worth upgrading or waiting even longer. First world problems...

SamuriHL

18th November 2020, 00:39

I don't believe I said a 4000 nit _OLED_ display! :p :D Yea, I'm looking forward to new display tech that keeps the fidelity we get from OLED but cranks up the NIT level. Ideally we want to get AWAY from tone mapping. I want to obsolete both DTM and DV. Obviously there's stuff over 4k nits but it's rare and I'm willing to allow the display to tone map "down" to 4k nits. ;)

Magik Mark

18th November 2020, 01:47

Guys,

Can you share resource on HDR Tone mapping using madvr? Most of the resources are converting HDR -> SDR.
But how about just plain HDR Tone mapping? I would like to compare it with LGs built in DTM

huhn

18th November 2020, 02:19

you have to understand that they are both in the end tone mapping to SDR.

so you if you are sure you can defeat the DTM just test it everything that applies to SDR applies to this too you simply have the issue of not really knowing how much of what you see is madVR tone mapping and how much is the TV.

VBB

18th November 2020, 02:32

Mark, that's what I was getting at above. I would also like to compare, but I need a better video card for that :(

SamuriHL and chros should be able to tell you more.

Oh, and I just realized you said DTM. If you want to stay true to the original, you want to turn that off. DTM is what LG does on top of its normal tone mapping, and it's all over the place.

SamuriHL

18th November 2020, 04:24

There's no easy "set this and you're done" kind of thing with these beta builds. And I can tell you, everyone has their own preferences and what they're going after. It may seem like it'd be "easy" to set it for HDR output but it can be challenging. I know some people post their settings. Take a look at NeoXP's settings and maybe start there, and tweak it to your liking.

Magik Mark

18th November 2020, 05:45

I'm now using NeoXP's. It looks really good. May I ask your preferred setting?

NeoXp's uses SDR feature of my OLED with gamma 2.4. Image a little bit dim. Is there a way to make it 2.2 using madvr tone mapping?

VBB

18th November 2020, 06:07

Neo has a really old OLED, that's why he does it. You want to use HDR to HDR tone mapping with his settings, but nits adjusted to 700 and "send HDR meta data" checked. Correct me if I'm wrong, guys.

Magik Mark

18th November 2020, 09:46

You are right VBB!

Just curious, shall I leave "dynamic target nics" alone?

chros

18th November 2020, 12:02

Mark, that's what I was getting at above. I would also like to compare, but I need a better video card for that :(
I thought the 960 *can* do pixelshader HDR output, it's not that weaker than the 1060 6GB (and mine is underclocked!). Try out this:
- use d3d11-native
- use b114
- use these settings (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60025149) from b113 and let the new options alone in b114
- disable Contrast recovery, it consumes more power when it kicks in (e.g. NeoXP doesn't use it at all)
- enable "Compromise on HDR luminance ..."
- I use NGU Sharp Low for chroma with the above settings, you may have to go lower a bit

Lemme know if it helps or not, it should work with 2160p@23p!!! :)

It is pretty complicated to desaturate even at 700 nits indeed, especially when we're talking about 2000+ nits source content.
As I said in the avs thread, I used desat 1+2 with b113, I tested back in the day and sometimes it made a difference.

These days I'm too lazy to boot into Windows to use madVR though.. so Dolby Vision it is.
:D same here (https://forum.doom9.org/showthread.php?p=1927703#post1927703).

I am fairly sure there are three different tonemapping curves in the C8.
...
So how does this relate to madVR? Well it seems that when you use pixel shaders and output in HDR you are actually setting an MDL in the output with the "display peak luminance". I tracked down where the different code values switch.
...
Am I missing something here? I just think that capping the DPL to 700 is missing some of the potential for the top end highlights and for no reason since the TV will tonemap regardless.
Almost, but it's a bit more complicated than this: take a look at this (https://www.avsforum.com/threads/2018-lg-oled-calibration-and-user-settings-no-price-talk.2962814/post-58739264) if you haven't seen it yet :)
But normally you'd measure your HDR light output with a meter (and e.g. HCFR) with a 1% window (on a 65" screen) and that's what you set for DPN in madvr. And on 2018 Oleds the built-in hdr peak is set to 700 (unlike with >=2019 sets where you can adjust it).

SamuriHL

18th November 2020, 13:50

You are right VBB!

Just curious, shall I leave "dynamic target nics" alone?That's a setting you'll have to play with. In fact with the latest build I have to go readjust it myself. It impacts how bright the image looks so you want to check really bright scenes with it. Just adjust it til it looks good to you.

Sent from my SM-G975U using Tapatalk

VBB

18th November 2020, 21:47

Lemme know if it helps or not, it should work with 2160p@23p!!! :)
Thanks! I'll give it a shot tonight.

oldpainlesskodi

21st November 2020, 09:19

Hi,

Firstly, sorry if this question is well trodden, but being new to the LG OLED, thought I'd ask here,as a lot of boards out there are contradictory.

Re the Color Gamut option - there are wide, extended and auto options.

Is it best left to auto for both SDR and HDR playback? or wide for HDR and extended for SDR?

Perhaps someone can enlighten me on what these options "really do"

Thanks

UPDATE - figured it out myself through testing.

VBB

21st November 2020, 21:23

For normal use, you want to leave them both on AUTO and set madVR to BT.709 for SDR. For LUT use, you want to set SDR to WIDE and set madVR to DCI-P3, just in case (it is unclear at the moment how that option interacts with the LUTs, at least in the latest beta). In PC mode, at least on the older models, the gamut is locked to WIDE, so you'll have to set madVR to DCI-P3, even if you don't use a LUT.

oldpainlesskodi

22nd November 2020, 09:44

Thanks.

Magik Mark

23rd November 2020, 05:03

Guys, Do you happen to know an ICC Profile for LG Oleds?

bombadilio

23rd November 2020, 09:16

mark0077

23rd November 2020, 23:23

Thanks! Would love your input Mark, SVP has an update that works with AVISynth + MadVR. I just can't seem to figure out how to tell if SVP is activated when using MadVR.

Do you know how to check? So far it seems to be working.

I am currently using:
Asus Rampage V X99 + 5930 @4.4ghs | EVGA FTW3 3090 | Creative Sonic Carrier Dolby Atmos | LG 55"CX | MPC-BE x64 -Madvr / AviSynth / SVP

Yeah I can try to help, but it might be best for the SVP forum. If you right click on the SVP menu and make sure it's set to show it's OSD if you want to be sure it's active. Also clicking on the tray icon while content is playing will show whether it's active or not (if your like me and don't want OSD). I just got the latest avisynth filter and enabled that in MPC external filter list, and disabled ffdshow. SVP on the fly injects it's scripts into the running avisynth filter instance while it's running.

PetitDragon

25th November 2020, 22:43

I just got the latest avisynth filter and enabled that in MPC external filter list, and disabled ffdshow.

hi mark,

could you post the link of that latest avisynth filter for me. TIA! :thanks:

mark0077

26th November 2020, 00:33

hi mark,

could you post the link of that latest avisynth filter for me. TIA! :thanks:

Sure. This is the GitHub page. https://github.com/CrendKing/avisynth_filter
This is the discussion thread on the SVP forum also https://www.svp-team.com/forum/viewtopic.php?id=5877&p=3

VBB

30th November 2020, 23:27

Feels like we haven't talked about smooth motion in a while :D

SamuriHL

30th November 2020, 23:37

All fixed up in the latest betas. Have at it. :D

chros

30th November 2020, 23:46

Feels like we haven't talked about smooth motion in a while :D
Screw SM, I don't like it :) But you forgot to reply: how's your 960 dealing with hdr pixelshader? :)

VBB

30th November 2020, 23:48

Oh hai! Yes, I know. I've actually switched back to SM and non-PC mode LMAO. Been re-watching Westworld in UHD, and I just couldn't handle the banding anymore after torturing myself through season one :)

Didn't forget. Just haven't had the willpower to test it LOL

chros

1st December 2020, 00:08

No worries. But WW is DV, don't you respect the director's intent?! :D

VBB

2nd December 2020, 00:51

It's ridiculous that after more than three years, I'm still switching back and forth between PC mode and normal mode. I can say with absolute certainty that for things like motion and lip sync PC mode is the clear winner. For all things SDR as well. I'm still torn on HDR, and I find myself comparing certain scenes where the banding is more obvious. Those scenes all have something in common: they are dark and have lots of saturated reds in them. It's also highly content-specific: Throughout season one (a 2K upscale) of Westworld I clearly noticed banding, and not just in dark and red scenes. Last night, I finished season two (native 4K), and except for the last episode, which had dark and deeply red scenes, I didn't notice any obvious banding at all.

Ideally, I would want to use non-PC mode for HDR, but the problem I'm having there is that motion in that mode is not smooth without Real Cinema, and enabling RC adds a crazy video delay. PC mode simply has the purest form of motion this TV can produce. And not to bring up smooth motion again, but I keep going back and forth on that as well. I realize I sound like a major flip-flopper, but I now tend to not like it as much as I used to. The blur is real at 60Hz :(

Asmodian, could you please re-assure us that all is good now in C9/CX land with your HDMI 2.1 10-bit RGB 120Hz decadence? :D

QBhd

2nd December 2020, 06:53

I settled on HDMI mode and TruMotion of De-Judder 2 and De-Blur 0.... HDR banding is a total deal breaker and the benefits of PC Mode do not even come close to making up that gap.

QB

VBB

2nd December 2020, 09:50

Thanks for the input, QBhd. After briefly trying that tonight and adjusting my Denon again (sigh), it might just make me come full circle.

huhn

2nd December 2020, 10:09

eARC to get the AVR out of the video pipeline can not come fast enough right?

Asmodian

2nd December 2020, 12:17

Asmodian, could you please re-assure us that all is good now in C9/CX land with your HDMI 2.1 10-bit RGB 120Hz decadence? :D

I have decided on 10 bit YCbCr 444 limited at 120 Hz. I cannot get a good full range calibration with the Calman integration. I dither to 8 bit with madVR but sending 10 bit from the GPU is not any worse, maybe even slight better.

I am very happy with it in this mode, smooth motion at 120 Hz is great and with the newest firmware on the TV (03.11.30) everything seem to be working properly.

It is a little odd to feel like I don't need any new technology for my display. I have been waiting for HDMI 2.1 and VRR for so long actually having it still doesn't feel quite real. :)

HDR banding is a total deal breaker and the benefits of PC Mode do not even come close to making up that gap.

This is the one upgrade for the CX over the C9. LG fixed HDR in PC mode.

bkrieger

2nd December 2020, 12:29

https://m.imgur.com/gallery/sXBmm4uHello,
I was hoping someone can help me. I have been getting a lot of dropped and repeated frames when watching HDR video. I’ve read this whole thread as well as other forums trying to figure it out and the correct settings. I have an LG CX 65, GTX 980 TI, soon to upgrade to an RTX 3080 TI, Ryzen 3700X, bit streaming with a Denon X4700Hhttps://imgur.com/gallery/sXBmm4u

Nvidia Control Panel RGB Limited, TV Blacks on Low non PC Mode

Here are my settings. Any help would be appreciated.

Thanks

bkrieger

2nd December 2020, 12:32

https://m.imgur.com/gallery/sXBmm4u

The images did not attach to my last post

https://m.imgur.com/gallery/sXBmm4u

https://share.icloud.com/photos/0u-MYKVlAEBt3je898DXFj1-Q

Asmodian

2nd December 2020, 12:49

quietvoid

2nd December 2020, 14:21

Reposting QBhd's post (thank you!) from AVS: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60296648
Very relevant, if some here have not seen it.
Anything lower than 850 target seems to roll off much earlier..

VBB

2nd December 2020, 17:52

+1 for QBhd's post. Saw it last night and thought, damn, that must have taken some time ;)

@As - Thanks for the update! Gives me some hope :)

@huhn - Can't tell if you're being sarcastic or not, but I want to say yes LOL

@bkrieger - Don't think anyone has mentioned it, but the main problem with the 980 is that it doesn't have a hardware decoder for HEVC. That's why you're having issues with HDR, but probably not with SDR.

QBhd

2nd December 2020, 18:39

Thanks for linking that here quietvoid... I was thinking that I should do that tonight... As for time, it really wasn't too bad. The i1Display Pro Plus is fairly fast and I use the Mehanik "Combined Hybrid Delay" files which line up almost perfectly with the timing of the measurements... only the 0% needs a pause of about 3 seconds. I use my ATi All-in-Wonder Pro remote and set the mouse pointer on the "Ok" button and just click every few seconds :P

As for interpreting the results for us OLED users... What I think is happening is that the DPL setting in madVR does not reflect the true nature of OLED. DPL in terms of a projector should not matter what size window is used. A 1% window should give the same results as a 100% window in terms of Peak Luminance. But as we know, that is not the case with OLED. If we go to RTings and look at what they measured, a 2% Peak is 979 (clearly not a sustained value). I think if we set DPL to anything lower than this we will lose those intense specular highlights and even lose a small chunk of 10% windows luminance (DPL 700 measured 724, while it is clear that the panel can go up to 761+).

What I think the value of Pixel Shader Tone Mapping for us is... all file types (different MDL/Max CLL/Max FALL) will be displayed the same. If you noticed some values were repeated in my test and highlighted red... those were using a 1000/1000/400 file, while all the rest were using 4000/4000/1000... and the results were exactly the same for those duplicates. The TV is going to tone map regardless what we do, but at least we can have madVR do all the heavy lifting of handling various HDR file types and send the TV something that is consistent.

QB

SamuriHL

2nd December 2020, 19:13

Yea, I agree with your assessment. If you can measure your display, that's what you should use for real display nits. The argument about 700, at least on the C8, is that it appears that above that the tv uses a more aggressive curve. We've no way to set the damn thing like later models do. So even though my display, depending on how I calibrate it, can hit close to 800 nits, I've set it to 700 to avoid the TV from going to a more aggressive curve. There may be a reason, though that the UB820 tone maps to 1000. Perhaps more experimentation is warranted given what you've shown.

oldpainlesskodi

2nd December 2020, 19:20

Sorry if this is a dumb question, but, certainly with the 2020 models, what are the "real" practical benefits of using Madvr to tone map? I have to say, straight passthrough is amazing.

Asking for a friend, of course,...

SamuriHL

2nd December 2020, 19:59

It's a question of static metadata tone mapping vs dynamic frame measuring tone mapping. It really can be remarkable what madvr brings to the table compared to static tone mapping solutions. If you've seen dolby vision on your display vs regular passthrough hdr...yea, it's like that.

oldpainlesskodi

2nd December 2020, 20:18

QBhd

2nd December 2020, 20:18

@SamuriHL... I would argue that at 700 DPL the C8 is far more aggressive in the tone mapping. Look at the curves between 501-800. They are all different and they all show a lot less maximum brightness. Whereas the curves again track better from 850-1200.... The 1100 and 1200 data points were just to show that nothing changes above 1000.

I think the C8 wants 1000/1000/400 content... Most movies have this as the metadata and the C8 tracks this very well. It's when we start going above and below this that curves start moving around.

I will be doing more research very soon...

QB

SamuriHL

2nd December 2020, 20:32

Which is why I want to try 1000 again and see how well that goes. It may be a better setting for us and maybe we can turn off some of the more aggressive settings in madvr while using it. (I'm thinking sky detection, shadow recovery, highlight recovery, etc). But we'd have to see how that works with an obnoxiously high NIT movie. Most movies barely go over 1000 or so anyway. But I agree, I think there's a real reason why Panny chose 1000 for OLED tone mapping.

SamuriHL

2nd December 2020, 20:38

hmm, tried it, (700 nits etc) I'm not coming at this from a confrontational perspective, but I really don't see it. Maybe its just meant for low nits projectors after all?

Emperors new clothes an all.

All good. Use what looks good to your eyes. It's going to be different for everyone. I have seen cases where madvr made a marked difference which is why i continue to use it. And we may be on to improving it even further. Question for you in your setup. Do you have LG's dynamic tonemapping turned on or off?

oldpainlesskodi

2nd December 2020, 20:47

I understand the reason for the question - it's on.

Your OLED is from 2019?

SamuriHL

2nd December 2020, 20:57

No I have a C8 which is 2018. There's a lot LG changed in the last couple models since mine.

As for dynamic tone mapping of the LG, that's not something they've improved. And that is also the reason you don't see a lot of difference between madvr and passthrough. As long as you're happy with that, it's good. I and others here are absolutely not.

oldpainlesskodi

2nd December 2020, 21:07

Ok, that's interesting...mine is a 2020 model....maybe they improved?

SamuriHL

2nd December 2020, 21:25

It's possible I suppose. But on the 2018 models, it messes with brightness in the mid SDR range. And other undesired brightness tweaks. Some people like the added brightness. Others like myself do not.

bkrieger

2nd December 2020, 21:32

Your settings are simply too high. You need to get your rendering time lower.

average stats
rendering 51.21ms
present 0.30ms

Both have to add up to about 40 ms or less for 24 fps video. You have to turn off some options, either disable some sharpening or set chroma upscaling to jinc or spline3 or both.

Edit: actually, your sharpening isn't active. Set chroma upscaling to Jinc. Jinc is the highest quality "pure math" scaler and it is much faster than NGU. I don't really like NGU low but you could also try that. Super-xbr or some of the other fancy options are also faster than NGU medium, you could try them.

Cubic (Bicubic60) is great for chroma upscaling, and super fast.

Or a faster GPU, of course. ;)

Thank You. I will try those options and settings. I had thought that NGU was the best to use and that was what I was using before I upgraded from my Panasonic 1080P Plasma to this LG CX. I do plan on upgrading to an RTX 3080 TI when released and in stock early next year. Couple more questions-

I had read that under exclusive mode under present several frames in advance, it has said to change to 1 and if any errors change to 3. Set at 1 I was getting glitches. Should I leave this at 8, or what number should this be?

Under general settings I have both GPU and CPU Queue set at 24. Is this correct? Also should present a frame for every VSync be turned on? I have Vsync and triple buffering turned on in the Nvidia Control Panel

Also in the Nvidia control panel should I leave color settings set to RGB and limited and blacks in the tv settings set to low? I had read that if RGB is set to full blacks need to be set to high

Under calibration settings should I leave it at BT 2020 or should it be BT 709?

Finally, under HDR settings is it ok to pass through to the TV or do I need to use tone mapping? The tone mapping settings seem complicated and I’m not sure what to set the settings at.

Thank You. I will try to play either the settings more tonight

QBhd

2nd December 2020, 22:00

Just to clarify again... The data I gathered was with 4000/4000/1000 files... with few points of 1000/1000/400 to check if there is a difference, there was no difference. This is our benefit.

I plan on eventually doing research into each of the settings I have disabled... But that is further down the road and I have other things in mind for the next session... Maybe tonight...

QB

SamuriHL

2nd December 2020, 22:11

Looking forward to seeing your results. I plan on doing some playing around tonight, as well. You've now got my curiosity peaked. :)

VBB

2nd December 2020, 22:18

I had read that under exclusive mode under present several frames in advance, it has said to change to 1 and if any errors change to 3. Set at 1 I was getting glitches. Should I leave this at 8, or what number should this be?

Default is fine here, or lower down to 3, which might be beneficial. Setting this to 1 is not necessary with the 9-series.

Under general settings I have both GPU and CPU Queue set at 24. Is this correct? Also should present a frame for every VSync be turned on? I have Vsync and triple buffering turned on in the Nvidia Control Panel

Leave the queues at default (16/8). Lower might be OK, higher is unnecessary.

Also in the Nvidia control panel should I leave color settings set to RGB and limited and blacks in the tv settings set to low? I had read that if RGB is set to full blacks need to be set to high

You definitely want RGB full there. Then it's up to what you do with the black levels on the TV. If all you do with your PC is watch media, I would suggest you set the TV to low. This will crush black levels for everything else, but playback will be fine. Make sure to set madVR to 16-235. If you use your PC for everything, then set black level to high and madVR to 0-255.

Under calibration settings should I leave it at BT 2020 or should it be BT 709?

If you're using PC mode with the TV, then the gamut is locked to wide (on my C7 it is), so you want to set madVR to DCI-P3 in this case. If you're using normal HDMI mode, set it to auto on the TV, and set madVR to BT.709. Make sure to also set madVR's gamma to match whatever you use with the LG.

Finally, under HDR settings is it ok to pass through to the TV or do I need to use tone mapping? The tone mapping settings seem complicated and I’m not sure what to set the settings at.

With the 980 in particular, you don't have a choice to use tonemapping, really. Like I said above, the 980 cannot decode HEVC in hardware, so you see a big hit in performance just there. I suggest you use straight pass-through and then tweak the chroma upscaling to whatever you can get away with.

See my responses in bold :)

bkrieger

2nd December 2020, 22:56

See my responses in bold :)

Thank You very much for all of your detailed responses. I mainly use my Pc for media as well as playing video games. The reason I had a question on black levels is I did have it on low with RGB Full, but I had read on another forum that if you set RGB To full Black levels should be set to high, and if you have RGB on limited then set the black levels to high. Is it stillOK to leave RGB on full and blacks on low? I have my PS5 source set to black levels on low as that is what is recommended.

Should VSYNC be turned on in the Nvidia control panel as well as triple buffering?

I have my TV set to HDMI mode as I had read mixed results with using PC mode. Is this correct?

Thanks again

VBB

2nd December 2020, 23:40

I leave all Nvidia 3D settings on default these days. I don't even go to the advanced settings anymore. So, VSync on "let the app decide" is just fine.

For black levels, like I said, if you have mixed usage with the PC, best to use RGB full -> madVR full -> TV high.

Normal HDMI is fine for most people.

Asmodian

2nd December 2020, 23:42

On the CX PC mode is very similar to HDMI mode, it is much less of an issue than it was with the C9 and earlier. The other modes also have much higher latency, so if you do use it for games I suggest keeping it in PC mode. I always use PC mode on my CX and I am quite happy with it, no downsides v.s. HDMI mode. I have calibrated my PC mode for SDR, with a BT.709 gamut, so I am not using it in its out of the box configuration.

Don't use VBB's suggestion of a mismatch of levels between GPU and TV. Since you game you cannot rely on madVR to always fix the black levels. VBB's suggestion for RGB full with TV set to low only works if you care about NOTHING outside of madVR.

For general use I get best results sending my CX YCbCr 444 limited instead of RGB limited. YCbCr is normally limited and RGB is normally full range. There is no reason to use ever RGB limited in my opinion, YCbCr limited is the better option in all cases. At best RGB limited is the same as YCbCr limited.

Set madVR to full, the GPU to YCbCr 444, and the TV to low.

For madVR leave the Vertical sync and Triple buffering options on their defaults, "Use the 3D application setting" and "Off" respectively.

VBB

2nd December 2020, 23:47

Asmodian

3rd December 2020, 00:20

Yeah, the PC v.s. HDMI mode is complex... test for yourself for best results.

My calibration might change it as well, and I never tested it uncalibrated. :o

SamuriHL

3rd December 2020, 00:24

chros

3rd December 2020, 01:59

As, you should mention, though, that PC mode is only similar to normal mode when sending an HDMI 2.1 signal. So, basically anything higher than RGB 8-bit @60Hz, because up until that, the TV still uses HDMI 2.0 processing, and will show the same banding we've always had in PC mode. I'm putting this here because anyone with a HDMI 2.0 or lower video card won't be able to benefit.

EDIT: It's possible this only applies to the C9, and not the CX.
I highly doubt this! :) The same SoC has to process 5x times data, how would this be possible? :) Plus we alredy saw that disabling certain image processing is a design decesion in PC mode by LG, just like player-led DoVi in mormal mode.
I think the best you can do is buy the TV then test the crap out of it and if it is similar to B7 then send it back :)

VBB

3rd December 2020, 02:18

Trust me, I found it hard to believe as well when I first read about it, but it's been proven and talked about ad nauseum over at AVS. It shows that LG can get it right, just not 100%, unfortunately. Us owners of the older models are SOL, of course ;)

QBhd

3rd December 2020, 02:42

My C8 is due for a new calibration. I'm just lazy. LOL I will likely do so before LOTR and Hobbit viewings though.

And I just changed madvr to 1000 real display nits. It does seem to make a difference but I need to do WAY more testing. It makes no sense for it to be aggressive with tone mapping under 700 nits but nothing LG does makes sense so....

Of course it makes sense... If we assume the display can do close to 1000 nits for short small peak bursts (specular highlights), then it makes sense that the display will tone map to maintain that. Using a 700 DPL in madVR is tone mapping to include only up to 700... Then the display wants to take that low nit input and tone map to what it thinks it should be, the 1000 assumption above...

QB

bkrieger

3rd December 2020, 02:57

Thank You. I will play with the settings and see what works, as well as PC Mode. The reason I had switched to RGB in the Nvidia Control panel is when I had it on YCbCr 444 I was getting bad lag and stuttering when trying to watch a movie. I’ll now lower the MADVR settings to see if that helps. I know I need to upgrade to a 2.1 GPU to have a better experience.

I appreciate all the help and advice

SamuriHL

3rd December 2020, 03:21

Of course it makes sense... If we assume the display can do close to 1000 nits for short small peak bursts (specular highlights), then it makes sense that the display will tone map to maintain that. Using a 700 DPL in madVR is tone mapping to include only up to 700... Then the display wants to take that low nit input and tone map to what it thinks it should be, the 1000 assumption above...

QB

Yea this seems to be holding true. I'm watching Rise of the Planet of the Apes right now at 1000 and it looks very good. I have highlight recovery turned off. Actually I have a lot of options turned off. This is interesting.

QBhd

3rd December 2020, 09:11

Okay so here is some more data...

Originally I had wanted to use HCFR and madTPG to try and send different luminance levels and track how “pass-through” behaves. Unfortunately, I can’t seem to change the output levels... however we can change the metadata being sent to the display. I am not sure what metadata madVR’s pixel shaders send, but I did read on one post it is something like MDL=DPL, Max CLL=DPL, and Max FALL=DPL/3 (DPL being “Display Peak Luminance" of course)... so that is what I used starting at MDL 300 (HCFR won’t let you use a Max FALL less than 100).

So here is the Data:

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/Data2.jpg

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/All2.jpg

As you can see it is a jumbled mess of info... so let’s parse it a little and try and interpret the results. First, lets compare madVR passthough to HCFR/madTPG for 1000/1000/400 and 4000/4000/1000 files/metadata... Just to make sure the data gathered actually reflects what the display will be doing during pass through

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/HCFR-PT.compare.jpg

So that’s good, both track almost perfectly and I would call any difference a measurement deviation between runs. BTW, if you are wondering about the Y Target of 768... In this session I saw that value was the most common and the highest. It’s just there as reference. Also you can see the aggressive roll-off of the 4000 file/metadata.

So again I did the 500/501 transition and here is how that looks:

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/500-501.transition.jpg

So strange... If you just look at the 300-500 curves they track nicely (just like in the pixel shader data I did earlier) and then when you add the 800-1000 it also gets back on track:

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/300-500+800-1000.jpg

Again, I think this all points to avoiding a DPL in madVR between 501-850... clearly we will not be setting DPL below 500, so that leaves what? I am going to go with what Rtings measured, 979. I don’t have the tools or the desire to go into the service menu to properly measure an HDR Peak 2% Window.

QB

chros

3rd December 2020, 10:30

What's your peak nit when you only measure 1 100% white patch with 1% window? (you don't have to go to the service menu)
And what is the number 768?

chros

3rd December 2020, 10:41

QBhd

3rd December 2020, 11:11

What's your peak nit when you only measure 1 100% white patch with 1% window? (you don't have to go to the service menu)
And what is the number 768?

That's the thing... I have tried to measure a 1% window and it keeps climbing until some sort of ABL kicks in and then drops, then climbs again... Plus, I can get it higher with a 2% window, which doesn't make much sense... As I said, I think my tools are limiting my ability to get a good measure of Peak brightness.

As for the 768 Target... EOTF is a moving target based on the 100% grey scale value recorded when using a 10% window. If you look at the data table I included, the highest value was 768 (and most common). I then just take that and plop that value as an "override" in HCFR to generate an EOTF target curve based on 768. It's just a reference based on the display's ability

QB

EDIT: LOL... I just noticed my typo in the charts... Yarget=Target I think I had originally wanted to call it Y Target and screwed up somewhere :P

SamuriHL

3rd December 2020, 15:43

bkrieger

3rd December 2020, 16:05

As, you should mention, though, that PC mode is only similar to normal mode when sending an HDMI 2.1 signal. So, basically anything higher than RGB 8-bit @60Hz, because up until that, the TV still uses HDMI 2.0 processing, and will show the same banding we've always had in PC mode. I'm putting this here because anyone with a HDMI 2.0 or lower video card won't be able to benefit.

EDIT: It's possible this only applies to the C9, and not the CX.

So I lowered some settings in MADVR including changing from NGU to Jinc, changed the queue size to 16, 8 and set the tv to PC mode. I seem to have made it worse is now movies are unwatchable and freeze. When I change the GPU settings from YCBR 444, the screen turns green. Any other YCBR setting the stuttering had been worse than when on RGB. I’m going to mess with it more tonight and try to out some settings back where it was as at least movies were watchable.

I was going to wait for an RTX 3080 TI for the 20 GB VRAM but I may get a RTX 3080 10 GB if I fine one

QBhd

3rd December 2020, 18:30

That's some great data and really does re-enforce the "don't use 700" as real display nit argument. Set it to 1000 and call it a day. Yes, if you could measure it you can get it more accurate potentially but clearly the TV's curve is set up to handle it at 1000.

Every time I saw a post with someone's settings or they recommended 700-750 DPL in madVR I cringed... I just knew it was wrong. It never made any sense to me knowing how OLED works (not the tone mapping, but the way brightness varies based on size of the bright patch) and what madVR was probably doing. Short of actually being able to get a proper measure of true Peak brightness all we can really do is look at how the TV handles various inputs and how madVR interacts with this.

Yes this is tone mapping on top of tone mapping. But that can not be avoided. All panels tone map, they need to since none even come close to 10,000 nits. But what madVR tone mapping will give us is valuable nonetheless... It will allow us to play videos of dramatically different HDR mastering and be able to know the TV is not changing anything between them. madVR will do that work.

QB

SamuriHL

3rd December 2020, 19:01

That's the point I don't think people outside of our little world seem to get. Not sending HDR is a non-starter given the MAX we can get out of these panels is ~400 nits with SDR. And that's pushing it. No thanks. If we can get madvr to tone map it into a range that we know our TV won't butcher then I think we're good enough. The 1000 nit range seems fairly consistent. If madvr can take a 4000 nit movie and compress it into a 1000 nit curve that we can send to the OLED, I think we'll get very consistent (as much as possible with OLED anyway) results. I will be playing with the settings more but it's staying at 1000. Way back when we first started playing with madvr that's what I had it set to and everyone told me 700 was the max the tv could handle and that it would use more aggressive curves above that. It would appear that advice was wrong and we need to kill it with fire going forward.

VBB

3rd December 2020, 20:16

So I lowered some settings in MADVR including changing from NGU to Jinc, changed the queue size to 16, 8 and set the tv to PC mode. I seem to have made it worse is now movies are unwatchable and freeze. When I change the GPU settings from YCBR 444, the screen turns green. Any other YCBR setting the stuttering had been worse than when on RGB. I’m going to mess with it more tonight and try to out some settings back where it was as at least movies were watchable.

I was going to wait for an RTX 3080 TI for the 20 GB VRAM but I may get a RTX 3080 10 GB if I fine one

What do you have your LAV decoder options set to? Choice of color space should have zero impact on performance. I've experienced the green screen with YCbCr, so I would avoid that. Could be an issue with only the 9-series video cards and newer drivers.

quietvoid

3rd December 2020, 20:24

Thanks for more info. I'm still confused if this is relevant when using "dynamic target nits"?
I've had all of these enabled for a very long time as they made madVR tonemap to higher targets, increasing the highlight details and contrast.

I guess I'd have to test this all myself all over again.

SamuriHL

3rd December 2020, 20:31

I'm in the same boat of having to retest and retune everything. However, given the data he's shown us, obviously it will be worth the effort because it's clear that it doesn't track so well at 700. It tracks damn near perfect at 1000 so if we can tweak madvr to give us a perfect 1000 nit curve it should in theory give us the absolute best quality HDR picture we can get on these screens. In theory anyway.

QBhd

3rd December 2020, 22:22

I actually don't think Dynamic Target Nits is for us. I have tested it on a few videos and it just makes the image dull. The one scene that jumps out was the Netflix Breaking Bad movie at the beginning where they are outside... DTN just made it flat and dim with almost no Dynamic Range. I would guess (not tested yet) that Fire Tweaks and Highlight Recover are still good to use... Sky stuff maybe.. And we don't need Contrast Recovery since we have high peak and perfect Black.

QB

chros

3rd December 2020, 22:42

So basically you don't use anything :D I mean, at 700 nits and above there's no work to be done with the majority of contents.
But madvr can do some crazy things as well, I sometimes switch between passthrough and pixeshader and the latter can be darker than the former :)

The 700 nits came from the calibration gurus who stated that the built in curve highest peak is set to 700 with 2018 LGs (just like with 2019 models by default). And on top of this the B8 has 200 nits less than C8 so for me it's 700.

QBhd

3rd December 2020, 23:04

SamuriHL

3rd December 2020, 23:28

QBhd

4th December 2020, 00:41

I think we should separate two things. chros may be right and that 700 is the internal target. What I suggest we stop doing is equating the curve to the peak. From what I understand, for a projector it is equal. For OLED it is not. Maybe for Calibration we should shoot for a 700 curve... I don't know, I have not calibrated mine... All measurements were with default White Balance and CMS.

If I was to make an informed guess... I would think that the C8 is optimal with 1000/1000/400 content and tone maps to a 700 curve... If indeed 700 is correct. But that 700 is not Peak Luminance.

QB

SamuriHL

4th December 2020, 01:13

That seems logical. But right now we can't really separate them. I don't calibrate HDR or DV specifically on my C8 because it's pointless to even try. I do have my white point set in the service menu to D65, though, and I use that for everything, including my calibrated SDR mode. In any case, I still think Panasonic chose 1000 nits for OLED for a reason and right now I am very happy with what I'm seeing by setting that as my DPL. Yes, being able to set 700 nits and force the metadata to 1000/1000/400 probably would be best but we can't do that right now.

Back to DTN for a moment, I think we absolutely DO need it. I just tried a 10000 nit video without it and it was an abject mess. As soon as I turned it back on and set it to 150 things are back to what I expect.

QBhd

4th December 2020, 05:17

I am guessing that is a 10,000 Nit demo video? Anywhere I can get a copy?

QB

SamuriHL

4th December 2020, 05:52

It's from the spears and munsill uhd demo disc. So you'd have to buy it but it's a very useful disc.

Sent from my SM-G975U using Tapatalk

bkrieger

4th December 2020, 07:31

See my responses in bold :)

I was able to get it to play smooth by setting the GPU to RGB Full, and the LG TV to PC Mode. Just want to verify, should MADVR be set to BT 7.09 or BT 2020 for both media AMD gaming? Also what should the sharpen edges screenbe set at?

Thanks

VBB

4th December 2020, 07:50

chros

4th December 2020, 09:06

Again... That's probably a 10% Window... Not the Peak that madVR will spit out... Super small super bright highlights will be capped to 700 well below what the display is capable of. That's why the measured 10% Window is actually brighter than 700 (730ish iirc)... I would guess that is the TV trying to pump up brightness to what it thinks is best.

Yes, I agree. So 700 is maybe 10% window on C8, but 1% window on B8. Let's say B8 can only do 200 nits less. So, if 1% is 700 on B8 then highlights maxed out at 800? That means 1000 for C8, and you don't have to care about "wrong" metadata.

Possible problems:
- we don't know the exact highlight peak value, that effectively means lieing to madvr :)
- what about movies like The Meg with really high max and avgFall? (I don't remember the exact values now) This can trigger ABL then...

QBhd

4th December 2020, 12:03

I wouldn't say using 1000 nits is lying to madVR... RTings value of 979 for a 2% Window Peak is pretty damn close to 1000

And just for reference the B8 managed 722 in the same test on RTings.

I am not saying that RTings is the gospel for such info, but they do have a long track record and take their methodology seriously... Let's call them a reliable reference

It's from the spears and munsill uhd demo disc. So you'd have to buy it but it's a very useful disc.

I actually think I might have a copy of that...

QB

bkrieger

4th December 2020, 12:17

Is PC mode still locked to wide gamut? If so, you'll want to set madVR to DCI-P3, because that's the closest to native. Are you talking about the edge sharpening setting on the TV? I have it disabled where possible.

How do I know if PC Mode is locked to wide gamut?
No, the upscale refinement screen. Should sharpen edges, crispen edges, and thin edges be checked and if so what should the numbers be?

Under properties in Madvr leave it on PC Level 0-255 and 10 bit?

Also when playing a movie when I hit Control J it says glitch 1. Any way to figure out what is causing that and get rid of it?

Still getting dropped frames but better than it was. I put the Madvr settings on low, Chroma Upscaling, Image Upscaling and image downscaling all on cubic 60

Thanks

chros

4th December 2020, 15:00

I wouldn't say using 1000 nits is lying to madVR... RTings value of 979 for a 2% Window Peak is pretty damn close to 1000

And just for reference the B8 managed 722 in the same test on RTings.

I am not saying that RTings is the gospel for such info, but they do have a long track record and take their methodology seriously... Let's call them a reliable reference
I agree but there are panel (settings?) diferences since my B8 only reaches 700 in 1% window using PC mode Full RGB. Anyway, normally everyone should measure his.

QBhd

4th December 2020, 16:37

What tools do you use to measure the 1% window (meter, software, panel settings, etc.)? I can't get a reproduceable result with my tools...

QB

SamuriHL

4th December 2020, 16:52

I may try to see what I can do with colourspace this weekend. Using my raspberry pi as the tpg I can force HDR using my Integral 2 so in theory I should be able to test 1% without even needing to touch the service menu. Maybe. :)

(Of course, once i set all that up I'm going to want to recalibrate the damn thing for SDR...it's been several hundred hours since I last calibrated it)

QBhd

4th December 2020, 17:39

I put 6.2 hours per day on mine... It's probably already over 3800 hours.

BTW, just to make sure we are on the same page, I use Cinema preset

QB

SamuriHL

4th December 2020, 17:44

Which will still be different for me because I've calibrated the D65 white point in the SM and use that for all my modes. So even though I'm using cinema I'm using a D65 white point.

VBB

4th December 2020, 19:17

How do I know if PC Mode is locked to wide gamut?

In the TV settings, Picture -> Picture Mode Settings -> Expert Controls -> Color Gamut. If it's grayed out and set to wide, it's locked.

No, the upscale refinement screen. Should sharpen edges, crispen edges, and thin edges be checked and if so what should the numbers be?

Entirely a personal preference. You'll have to judge for yourself. I don't use any refinements.

Under properties in Madvr leave it on PC Level 0-255 and 10 bit?

If your video card is set to RGB full and the TV is set to black level high, then yes. 10-bit is up to you. I would recommend going with 8-bit due to reduced banding.

Also when playing a movie when I hit Control J it says glitch 1. Any way to figure out what is causing that and get rid of it?

When you first start playback, there will be drops/glitches. That's perfectly normal. Give it a few seconds, then hit CTRL+R to reset.

Still getting dropped frames but better than it was. I put the Madvr settings on low, Chroma Upscaling, Image Upscaling and image downscaling all on cubic 60

You could also try setting the LAV decoder to "none", which is software decoding using the CPU. Maybe that will improve performance in your case.

Thanks

Hope that helps :)

SamuriHL

4th December 2020, 23:14

I did a little metadata testing. This is on a movie I'll be watching tonight. I wanted to compare the HDR optimizer on my UB820 to what madvr is sending.

madvr:
LUM 1000/0.005 maxCLL/FALL 1000/333

ub820:
LUM 1000/0.005 maxCLL/FALL 1000/0

passthrough:
LUM 1000/0.005 maxCLL/FALL 0/0

Admittedly not super interesting on a movie that:
min: 0.0050 cd/m2, max: 1000 cd/m2

So I pulled out the good ole Spears and Munsill demo disc. That's got a variety of fun content to play with. I decided to play with the 4000 nit BT.2020 HDR10 demo video.

madvr:
LUM 1000/0.005 maxCLL/FALL 1000/333

ub820:
LUM 1000/0.005 maxCLL/FALL 1000/800

passthrough:
LUM 4000/0.005 maxCLL/FALL 4000/1616

A little more interesting than my movie I suppose:
Mastering display luminance : min: 0.0050 cd/m2, max: 4000 cd/m2
Maximum Content Light Level : 4000 cd/m2
Maximum Frame-Average Light Level : 1616 cd/m2

I'm not sure exactly what this tells us but clearly the UB820 is also screwing around with the metadata that it's passing to the OLED display during it's tone mapping algo, as well. It seems to be chasing a higher FALL than madvr does though. I forget how madvr calculates the FALL. Anyway, I thought I'd post this in case it gives anyone food for thought. I simply found it an interesting exercise.

QBhd

4th December 2020, 23:49

I think it was chros who recently mentioned that madVR outputs MDL=DPL, Max CLL=DPL, Max FALL=DPL/3

I don't think LG cares about Max FALL... But your findings again point to 1000/1000 (MDL/Max CLL) as being optimal metadata for these devices.

QB

SamuriHL

5th December 2020, 00:11

I would agree with that.

Also I just set up for calibration (what a pain in my case but anyway...will probably do SDR while I'm at it). I took several readings of L10 which I believe is a 1% window size in Colourspace with metadata sent as 1000/1000/400 and watched it go up to ~890 or so. It does start to fight with me if I leave it like that for too long and starts dropping again. I'm not sure if I'm exactly measuring this correctly, but it seems reasonable to me. Nonetheless, I would definitely argue that you are correct about specular highlights getting into the 900+ range. Which again, points to metadata 1000/1000 being ideal.

bkrieger

5th December 2020, 03:18

http://https://share.icloud.com/photos/0WNACTrK5UNBY1MAgkxd4b__QWhat do you have your LAV decoder options set to? Choice of color space should have zero impact on performance. I've experienced the green screen with YCbCr, so I would avoid that. Could be an issue with only the 9-series video cards and newer drivers.

I had seen an updated beta Madvr I believe it was version 113 that I downloaded the other day. I did not release but that is what caused videos to be unwatchable. I am now back to version 92.17 and is back to how it was before.https://share.icloud.com/photos/0lJX_jZdFr9dgd261fbInrNIwhttps://imgur.com/gallery/t7VanKo

Here are my Lav settings

https://imgur.com/gallery/t7VanKo

https://share.icloud.com/photos/0-9XJENHC41przubaXt8-mShQ

bkrieger

5th December 2020, 03:37

Hope that helps :)

The LG Is locked to Wide Gamut using PC Mode

Thank You for all the settings tips and advice. I will try those and see if it helps.

I appreciate it.

Thanks

chros

5th December 2020, 08:37

What tools do you use to measure the 1% window (meter, software, panel settings, etc.)? I can't get a reproduceable result with my tools...
What do you mean it is not reliable? +/-10 nits in hdr is not an issue. :)
I have the same tools as you, nothing special, I've listed in that B8 post :)
Obviously the panel is pretty unstable in HDR. I used the settings that are documented in that post, and even with that bfi setting, the grayscale patches heat up the display, let alone if you leave the 100% white there just for 30.seconds, the peak nit is rising slowly :)
BTW, just to make sure we are on the same page, I use Cinema preset
I used that as well.
Which will still be different for me because I've calibrated the D65 white point in the SM and use that for all my modes. So even though I'm using cinema I'm using a D65 white point.
Me too, but can't we set this in HDR10 Cinema preset? If we can then I don't think it matters much.
Let alone using 100/120 nits SDR D65 for 203 nit HDR diffuse white... :) How does this even work? :) I guess if LS/CS gurus don't have an issue with it then it's good to go, right?! :)
I think it was chros who recently mentioned that madVR outputs MDL=DPL, Max CLL=DPL, Max FALL=DPL/3

I don't think LG cares about Max FALL...
I asked madshi about the calculation around a year ago, before I wrote that B8 post.
Yep, dunno about maxFall, but maybe LG uses it for ABL algos, who knows :)
I took several readings of L10 which I believe is a 1% window size in Colourspace with metadata sent as 1000/1000/400 and watched it go up to ~890 or so.
That's about right with the C8 (+~200 than B8). I told you guys to measure your C8s back in the day! :)
What sad is that newer CX models has almost the same peak as my B8 :) The C8 is the brightest out of all so far (except for the special Panasonic editions, gz2000/etc).

QBhd

5th December 2020, 08:50

Hmm... I actually had to look that up... it seemed odd that the C9 and CX would be less bright than the C8... but yeah a quick peek at RTings does confirm that... Wow

Not that I didn't trust you... but I have always been a guy that fact checks ;)

QB

Asmodian

5th December 2020, 09:26

it seemed odd that the C9 and CX would be less bright than the C8... but yeah a quick peek at RTings does confirm that... Wow

I noticed the trend downward even when calibrating 100 nits SDR, moving from the C7 to C9 to CX, each one needed a slightly higher OLED light to hit 100 nits after calibration.

I assume it is due to concern over burn in.

QBhd

5th December 2020, 13:40

chros

5th December 2020, 16:21

Not that I didn't trust you... but I have always been a guy that fact checks ;)
:) No worries, I can also remember wrong.
I think you have to have some seriously bad viewing habits for burn-in to happen... I am approaching 3800 hours on my C8 (6.2 hrs/day) and I don't have any burn-in. Plus I think I like my SDR brighter than most (OLED Light is at 56)
Yes but LG is concerned about the average Joe not us :)

SamuriHL

5th December 2020, 17:54

Me too, but can't we set this in HDR10 Cinema preset? If we can then I don't think it matters much.
Let alone using 100/120 nits SDR D65 for 203 nit HDR diffuse white... :) How does this even work? :) I guess if LS/CS gurus don't have an issue with it then it's good to go, right?! :)

There's been a lot of debate about white point in HDR as you know. I'm not knowledgeable enough to argue the merits of either or. In any case, yes it can be set very easily because you just set the color temp to the one you calibrated. In fact, ConnecTEDDD recommends calibrating 2 different color temps in the SM, one for SDR and one for HDR. I've not done that yet but that's how you can get your 203 nit HDR white if you so choose.

That's about right with the C8 (+~200 than B8). I told you guys to measure your C8s back in the day! :)
What sad is that newer CX models has almost the same peak as my B8 :) The C8 is the brightest out of all so far (except for the special Panasonic editions, gz2000/etc).

I did measure it back in the day and it was lower. I was also using Calmann back then and attempting to calibrate HDR and DV. Which is futile on these panels. Simply calibrating white point is good enough for HDR and DV. Obviously I have a 3dlut for SDR and I do not touch the 1dlut at all (stupid LG bugs!). But yea, the C8 is quite the bright display which I love. I can see very clearly on some of my test content that by setting DPL to 700 we were indeed limiting the overall specular highlight brightness.

I think it's time for a LOTR viewing...

SamuriHL

5th December 2020, 17:55

I think you have to have some seriously bad viewing habits for burn-in to happen... I am approaching 3800 hours on my C8 (6.2 hrs/day) and I don't have any burn-in. Plus I think I like my SDR brighter than most (OLED Light is at 56)

QB

I come from plasma so burn-in avoidance is in my blood. And, uh, yea, that's quite bright for SDR. Mine's set at 30. :)

QBhd

5th December 2020, 20:08

My living room is not super bright, but it is very far from dim... RTings recommended 56 for bright room and when I finished testing various things it worked the best for my taste. I have a triple lamp in a far corner pointed at the wall and ceiling... I don't watch SDR movies anymore and for TV shows, OTA, and YouTube it's good enough... When watching an HDR movie I will turn one or two of the lights off... Dolby Vision is black out, and wow does it look amazing.

I too came from Plasma, so yes B-I avoidance is always in the back if my mind

QB

VBB

5th December 2020, 20:20

I don't watch SDR movies anymore

:eek::eek::eek: :D

Nothing wrong with a brighter SDR picture. I settled on 46 years ago.

SamuriHL

5th December 2020, 21:14

Blasphemers! :D LMAO! No I get it. I have a very bright room in the day time but I do most of my watching at night.

I gotta say, Fellowship of the Ring is damn impressive. Yes, I could have watched it in DV but I wanted to see what HDR would do with my new settings.

chros

6th December 2020, 09:57

Yes, I could have watched it in DV but I wanted to see what HDR would do with my new settings.
Herecy! You don't respect the director's intent either! :D

SamuriHL

6th December 2020, 16:53

Well if they'd fix madvr I would! :p :D

SamuriHL

7th December 2020, 17:22

I did some recalibration this weekend. I created a white point for 203 nit to use for HDR. I need to do some more testing on that. At first glance the results are interesting. I'm not sure they're correct, though.

I also changed my SDR calibration to DCI P3 D65. I've come to the realization that this display is 95% DCI P3 and that all of the source I play on this display are capable of sending that. Some quick tests with madvr set to DCI P3 calibrated display and the bt.2020 checkbox set looks really good. I'll be playing around more this week but so far this has been fun.

QBhd

7th December 2020, 18:23

Is that 203 Nit white point really a good idea? When I Google HDR white point it still says 100ish (the 93.blahblah of SDR)... I am still in the process of learning calibration, that's why I just measured for now.

QB

SamuriHL

7th December 2020, 18:45

Probably not, no. There's a whole huge debate on this topic. Most UHD's are mastered around 100 nit white point while supposedly newer ones are using a 203 nit white point. Maybe. I did it just to see what the result would be. It's super easy for me to switch between them (and because I don't create a 3d LUT for HDR it hardly matters) for comparison. Which is why I did it.

Do I think it's a "good idea?" No, not really.

QBhd

8th December 2020, 04:16

Okay... so I pulled out my S&M UHD disc... and that 10,000 Nit demo is interesting to play around with DTN. I am not sure DTN is really the best way to describe this feature of madVR. This demo has tons of excellent scenes with varying FALL and CLL. So when you get to the horses in the snow it is very evident that DTN is needed. So I did a test of DTN=1 and DTN=500 (the highest you can set it) I have two profiles for HDR and set-up a short cut to switch between the two. 500 made the horse scene come to life, 1 basically did nothing.... anyway, I used this technique to find out that all you need to make the horse scene awesome was a DTN of 10... okay that's interesting... but the scene a few before that with the sun shining through a valley on the top of the mountains where the peaks are bright and the rocks are shadowed a little... this was still overblown and the contrast was not there, so I did the switching back and forth for a bit and found that 50 DTN brought that scene to life (without changing the horses).

This value seemed good since it was not dimming the low nit scenes, yet giving those scenes that needed it room to show the range... but then I got to that scene with the buffalo on a yellowish field... the yellow is still overblown here... so 75 DTN... not perfect, but it maintains excellent range on the other two scenes.

I think Dynamic Target Nits is just a wrong description... it's more like it is giving more room to the dynamic range, and it's not a nit value per say but more like a percentage... especially if you pull up the detailed HDR info on the OSD.

Anyway, I did all this with no compr limit set at 0, ... blah... let's just do a SS:
https://bn1301files.storage.live.com/y4mof0PbSxmakrmgKbBu81vG7n6R9CIvTdf5ONlaPScot2UlnVBROtuOV9Myljy5xzOdVx7p939OJRyx1wrYF5Hv0TAv3Db3ABVpSvHR3aWW9iuJ6ZCNTTbsLEa8vMfDC2mRVEsHHEfeYQfmA8BgbkIjP45TaZl2lDC-jjKFjADetvOmpCUMFKa_QebpF1NJ0kD?width=826&height=410&cropmode=none

So if I get a chance later tonight, I will explore what no compr limit actually does.

QB

SamuriHL

8th December 2020, 05:24

Yes, it IS a percentage. It allows some scenes to go above your DPL. Which is very useful for us. That's why I said we need it. I truly believe that if we can set the DPL accurately, metadata to 1000/1000/400, and dial in the DTN we'll be in heaven with these displays. I haven't had time to redial it in. But those scenes, yes, that's why I use them. To evaluate madvr tone mapping. Because at 10000 nits, you're getting some serious compression. And because he includes lower nit versions, you can compare to what it should look like. This is where the fun begins. :)

QBhd

8th December 2020, 18:35

I've been thinking about the behavior I saw paired with the measurements I have already taken and I don't think it's about going above the DPL. It's fitting more range into the DPL window. If you have DTN off (or set to 1, which is basically off) anything above 1000 (the current theory's best MDL for the C8, and probably really close to it's actual Peak Luminance) is basically clipped and squeezed at 1000. With a 10k source that is almost all white is displayed the same and it looks like hot garbage, literally... but the scenes that are below 1k all look great. If we slowly increase DTN, more range is allowed into out 1000 nit window and as such we see more detail in those blown out whites... however there is a cost, those darker scenes start to dim since they are now being squeezed too. With my quick and dirty testing last night I came to the conclusion we do need some, but we also don't need much. For example the horses in the snow only needed 10, anything above 10 had no perceptible effect on that scene (I think it was 10, if not, it was shockingly low to what I expected). There is a balance that must be found. Plus there is the "no compr limit" that needs to be explored (next on my to do list) to see what that is actually doing with a 1000 nit "window"... I am pretty sure what that will be doing, but the engineer in me needs to test and verify to be sure... and who knows, maybe it too does things that are not entirely evident in the name.

DTN is a terrible name... I always thought it was a target nit that madVR would shoot for. Like we have 1000 DPL set and if we set DTN to 200 we would have an image that is around 200 nits. Since I like a brighter SDR, it made sense to "target" a brighter overall HDR... but that is not really what I see with a 200 DTN. It's more like there is 200% more range to be used, and this really does a number on the lower nit scenes... anyway, terrible name :P

QB

SamuriHL

8th December 2020, 19:49

Agree on the name. It has confused nearly everyone.

My understanding (which very well could be wrong) is that DTN will allow a frame to go above DPL in some cases. My thoughts on exploiting this is to set things up like this:

DPL: Measure 10% window nit level (e.g. 750 or whatever)
metadata: 1000/1000/400
DTN: tbd but enough to get us to 1000 nit on 10000 nit scenes

If we can have this so that highlights can go up to the theoretical 1000 nits that our chosen curve allows (and we know the 1% window on the C8 can nearly accommodate) then madvr should tone map everything to the DPL while allowing highlights to go above it, which is exactly what we're trying to achieve. But this is all complete theory right now. It's just my wish.

quietvoid

8th December 2020, 20:28

I don't think DTN is supposed to be affecting scenes with a peak under the DPL, at least from previous testing.
Increasing it does increase contrast and highlight detail, at the cost of making the rest of the picture darker.
In the horses in the snow scene, this is obvious when comparing to passthrough. The horses lose all their color and just look dark.

Maybe setting a high no compresssion limit would help avoid making things too dark in those very high brightness scenes, never tested.

SamuriHL

8th December 2020, 22:18

I've not tried that, either. It would be a good test to try.

chros

8th December 2020, 22:59

QBhd, you are right about what DTN does. I use 55 (I don't remember why :) ), it makes the image a bit brighter when it's in effect.

But my biggest question is this: what does or how does behave the LG internal 700 nits target?! I can't test it, since my peak is around that.
Maybe you guys can take a look, with content between 700 and 1000 nits and setting madvr dpl to 700 and 1000. (but 700 results in brighter image on LG). Because if you set 1000 but LG is clipping at 700...

SamuriHL

8th December 2020, 23:28

That's why I very badly want to be able to set DPL to a measured real display nits and then use the metadata to force it into the proper tone curve. Technically speaking I can PROBABLY do this already. I may try here in a minute.

For me I'm really liking what I'm seeing at the moment with:

https://i.ibb.co/Z6y8L9f/madvr-2020-12-08-17-23-54.png (https://ibb.co/qMbsy6D)

chros

8th December 2020, 23:31

And guys please take a look at this post (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60278979)!
I wrote about this 2 times already and literally no one gave the slightest FAQ :) That's a big bug especially for us.
If you agree then quote that message and write madshi (preferably before the next beta) otherwise it will never be fixed :)

It should be only 2 "if"s:
- don't do any clipping if measured frame peak is lower than DPL
- if it's higher then only reduce frame peak nit to DPL (and not below it!)

quietvoid

8th December 2020, 23:43

Dynamic clipping is essentially a way to counter low nits titles in SDR, they usually end up flat otherwise.
Examples IIRC are Game of Thrones S01, Murder on the Orient Express (2017).
I think the same overall brightness uptick can be applied to HDR, however it goes against the coded PQ intended value.

It also helps in weird content like Mad Max and that Harry Potter film, where you would have grey-ish highlights if they were not clipped.

chros

8th December 2020, 23:45

I think it's useful for us as well (Mad Max Fury road, etc).
But even if it isn't, a bug is a bug! :)

quietvoid

8th December 2020, 23:47

I'm not sure if it is a bug, though I have no idea what the algorithm is and when it applies.
It could apply to both scenes lower and higher than DPL for all I know.

SamuriHL

8th December 2020, 23:51

I bumped it for you. I have no idea if he'll actually look at it though. But hey, I tried. :D

We REALLY need the metadata override in madvr. I just created my own. The settings I posted above changed to 700 DPL. Then I changed the metadata to 1000/1000/400. Yup. That's what we want to be starting from!!

Also, setting the no compression to 100 definitely seems to help. You guys might want to play around with that.

QBhd

9th December 2020, 00:02

I'll be doing another measurement session tonight, I will be home in about 5 hours. I plan on measuring Pass-through, DTN off, DTN=1 NCL=0, DTN=500 NCL=0, DTN=1 NCL=1000, and DTN=500 NCL=1000.... And few points in between. If there are other points of interest let me know here.

QB

SamuriHL

9th December 2020, 00:20

Can you test the NCL at 100?

chros

9th December 2020, 00:37

I'm not sure if it is a bug, though I have no idea what the algorithm is and when it applies.
It could apply to both scenes lower and higher than DPL for all I know.
I'm pretty sure it is: the algo (implemented by Soulnight, btw what happened him?) purpose is (for pj users) to clip certain amount (%) of highlight to be able to compress less. Now, if nothing to compress then there's nothing to clip :)

@SamuriHL, cheers!

chros

9th December 2020, 00:56

We REALLY need the metadata override in madvr. I just created my own. The settings I posted above changed to 700 DPL. Then I changed the metadata to 1000/1000/400. Yup. That's what we want to be starting from!
Honestly, I don't think you guys with C8 need it.
I took a look again of my B8 measurements (https://www.avsforum.com/threads/2018-lg-oled-calibration-and-user-settings-no-price-talk.2962814/post-58739264). Take a look only those that are 1000 or less, the interesting part is:
- 300-500 and 840-1000: the curve is the same!!!
- 501-800 (?): LG does some crazy shit for whatever reason, and as it goes up it tries to correct itself slowly. As we can see 760 is still a bit brighter than the normal but it's really close to it.

Unfortunately I haven't measured at 800 :) (and straight above 1000) I don't know what the turning point is exactly (maybe 801? :)) but 840 is fine already.
And the measurements are correct: 840 and 960 have almost the same values!

And that's what QBhd results show as well.
So, that means for sure, that you can set DPL between 840 and 1000 and the curve will be the same. What the display does with rest is a good question, but probably only for me with 700 nits with 1% window :D

@QBhd: measure 1% window (if you could go below it then do so :)), please put the measured points and their values on the graph as well.

QBhd

9th December 2020, 01:08

1% window is almost useless info with the tools I have... I will do it, but the data is junk.

QB

SamuriHL

9th December 2020, 01:42

Honestly, I don't think you guys with C8 need it.
I took a look again of my B8 measurements (https://www.avsforum.com/threads/2018-lg-oled-calibration-and-user-settings-no-price-talk.2962814/post-58739264). Take a look only those that are 1000 or less, the interesting part is:
- 300-500 and 840-1000: the curve is the same!!!
- 501-800 (?): LG does some crazy shit for whatever reason, and as it goes up it tries to correct itself slowly. As we can see 760 is still a bit brighter than the normal but it's really close to it.

Unfortunately I haven't measured at 800 :) (and straight above 1000) I don't know what the turning point is exactly (maybe 801? :)) but 840 is fine already.
And the measurements are correct: 840 and 960 have almost the same values!

And that's what QBhd results show as well.
So, that means for sure, that you can set DPL between 840 and 1000 and the curve will be the same. What the display does with rest is a good question, but probably only for me with 700 nits with 1% window :D

@QBhd: measure 1% window (if you could go below it then do so :)), please put the measured points and their values on the graph as well.

You're missing the point I think. I'm aware that the tone curve between 501 and 800 is fubar. But the DPL falls in that range. What I REALLY want to do is get madvr to tone map to the actual DPL then use the metadata to force the C8 to follow a much nicer curve. In theory, if we get DTN to work properly for us the way I want, we should be able to allow specular highlights above DPL so we don't artificially lose brightness where we need it, but, everything else gets mapped to the DPL. IOW, *I WANT IT ALL* :D

QBhd

9th December 2020, 02:00

I still think you are mistaken in thinking that madVR DPL=OLED Peak Luminance. And because of this, you think madVR will output higher than DPL with DTN... Nothing about that makes any sense to me..

QB

SamuriHL

9th December 2020, 02:48

I'm not entirely sure. It's very difficult to go back into that thread to dig up what DTN is ACTUALLY doing.

from madshi:

"dynamic target nits". All it does is change your picked "display peak nits" value, depending on the properties of the frame.

He doesn't explicitly say it goes above the DPL. So I don't know.

chros

9th December 2020, 11:59

1% window is almost useless info with the tools I have... I will do it, but the data is junk.

I still don't understand why/how ... Take a look at my hcfr settings (https://www.avsforum.com/threads/2018-lg-oled-calibration-and-user-settings-no-price-talk.2962814/post-58739264) (including bfi, etc) (see the images of hcfr in the attachement section under the post), I got 99% repeatable results with it (and I used the same equipments as you do).
Yesterday night I had to double check the 840 and 960 result images because I thought maybe I uploaded the same image twice (the values are almost the same) :D

I still think you are mistaken in thinking that madVR DPL=OLED Peak Luminance. And because of this, you think madVR will output higher than DPL with DTN... Nothing about that makes any sense to me..
Yep, he doesn't understand what we meant :)

You're missing the point I think. I'm aware that the tone curve between 501 and 800 is fubar. But the DPL falls in that range. What I REALLY want to do is get madvr to tone map to the actual DPL then use the metadata to force the C8 to follow a much nicer curve. In theory, if we get DTN to work properly for us the way I want, we should be able to allow specular highlights above DPL so we don't artificially lose brightness where we need it, but, everything else gets mapped to the DPL.
It doesn't work like this :) Take a look at my response on avs (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60316236) and after read what I wrote above (https://forum.doom9.org/showpost.php?p=1930289&postcount=1155), it will be clear :)

mikahawkins1

9th December 2020, 16:39

Is there any way for optimizing chroma?

VBB

9th December 2020, 19:33

Optimizing chroma in what way? What's your gear?

QBhd

9th December 2020, 20:06

So that was a big fat nothing burger! But even then, the data still can be useful and conclusions can be made. So like Parts I & II let’s barf out the data:

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/Data3.jpg

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/All3.jpg

Let’s get rid of the Pass-Through and the 1%... they are just cluttering things up and not very useful.

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/All3.1.jpg

So I highlighted two sets of data and used arrows to show these two on the chart... Both are with DTN Off or set to 0 (It makes sense that would be off as well). That isn’t so odd in itself, but what makes this odd is that just a DTN=1 bounces the curves right back in line with all the others. What does this mean? Not sure, does it point to DTN being needed? Is it just an oddity? (although two sets each with DTN off or DTN=1 is consistent enough for me when everything else is nearly exactly the same).

Okay, so 20-point Grey Scale does nothing to show what either of these two settings do. I must say, that was not the results I was expecting. I used the extreme data points in the hopes of uncovering behavior. But I guess these types of settings really only show themselves in real world content.

So I loaded up the trusty S & M UHD disc again and loaded all the versions this time: SDR, 600, 1000, 2000, 4000, and 10000. I then setup my two HDR profiles with DTN=75 for both and NCL=0 on one and NCL=200 on the other. I know SamuriHL was focused on NCL=100, but I figured with these being subtle settings for us, I should set a bigger NCL to see what the difference is more easily. Well, I couldn’t see a single difference... no good... no bad... nothing, nada, zip, diddly squat! And that was on each version of the demo.

No compr limit.... it doesn’t seem to do anything for us

I did have one take away from last night’s session that was awesome... All the demo files looked exactly the same with madVR pixel shader tone mapping... well the SDR didn’t of course and the 600 nit version was perceptibly dimmer (not much, but noticeable). This is our benefit... making our TV think any type of HDR is the same.

QB

chros

9th December 2020, 22:27

The DTN off and 0 rssults are interesting, I agree that it's not a coincidence. Maybe It's a small bug in the logic that makes it deviate from the curve. Report it to madshi on avs he can fix it if he is interested in it and it is indeed a bug.
I also agree that these settings only play with real content. DTN e. g. depends on the frame FALL as well.

NCL: I haven't seen its effect either here. But it has to do something (if it is not broken :)) I use 100, Manni suggested it with his 120 DPL! :) The question is does it affect the rest of the settings, e. g. shadow recovery :)

It's just insane that we have to experiment things to find out how they work... That's why I told madshi that at this point we need a small doc about these settings, what their purpuse is and how they should behave by itself and interact with each other.
As usual, no reply at all...

quietvoid

9th December 2020, 23:13

I was confused at why the images are not showing, Firefox is blocking them because the SSL cert is bad.
I didn't really expect NCL to affect the peak brightness, just everything below the set value. So not sure these prove anything for me.

QBhd

9th December 2020, 23:44

That's why I went to actual content to test NCL... and I could not see a single instance where it did anything for us.

I am actually in the process of systematically going through ALL 71 scenes on the S & M 10k Demo trying to narrow down the perfect DTN... so far 75 looks great on most, but there are a few scenes where you can see 75 still hurts it ever so slightly... I know I had tested 50 earlier and it wasn't enough for some... so I am working on 62. And narrowed it down to the first half of the demo (it's a bit of a chore, but worth it)

For those wondering how I am doing this... I have two profiles setup each identical except differing DTN. I pause on each scene and toggle between the two. I have a remote that allows for buttons to be mapped to keyboard shortcuts... it's only a chore since I am going through scene by scene :P

QB

chros

10th December 2020, 00:22

:) That's what I call dedication! Thanks for your hard (!) work. Let us know what the outcome is.

QBhd

10th December 2020, 01:21

So DTN... We need it... but how much to use. Of the 71 scenes there are at least a dozen where ANY DTN is good. But the thing is, once you hit the sweet spot for that scene more does nothing. On the flip side there are at least a dozen scenes where ANY DTN has a negative effect (even if slightly and hard to see). So my starting point was 50 not being enough for the buffalo and 75 being the number I threw in there the other night to get that scene to awesomeness. I split the difference and went for 62 vs 75 with 62 winning the vast majority and it being a toss up for most of the rest... however, the buffalo still won at 75. Now I know what you are thinking, why go higher for just one scene? Well, it boils down to this: when more DTN is needed it is really noticeable, when it is too much it is so hard to tell. My final destination is yet again splitting the difference and going with 65.

Going higher than 65 will have zero benefit on the scenes that need it and for the scenes where it's too much it is very hard to tell it's too much.

QB

VBB

10th December 2020, 01:56

Video RAM, sure ;) Normal RAM, nope. Certainly not when going from 32 to 64.

chros

11th December 2020, 09:52

Thanks for your test, @QBhd!

So DTN... We need it... but how much to use. Of the 71 scenes there are at least a dozen where ANY DTN is good. But the thing is, once you hit the sweet spot for that scene more does nothing.
Interesting. Does it make the image dimmer if you use way higher (e.g. 200 DTN) than the sweet spot?

On the flip side there are at least a dozen scenes where ANY DTN has a negative effect (even if slightly and hard to see).
In what form?

So my starting point was 50 not being enough for the buffalo and 75 being the number I threw in there the other night to get that scene to awesomeness. I split the difference and went for 62 vs 75 with 62 winning the vast majority and it being a toss up for most of the rest... however, the buffalo still won at 75. Now I know what you are thinking, why go higher for just one scene? Well, it boils down to this: when more DTN is needed it is really noticeable, when it is too much it is so hard to tell. My final destination is yet again splitting the difference and going with 65.

Going higher than 65 will have zero benefit on the scenes that need it and for the scenes where it's too much it is very hard to tell it's too much.

Nice! What DPL did you use for these tests?
Of course SM sample is really extreme. Do you mind quickly go through these real samples (https://mega.nz/folder/n5dzgQbJ#A99MVeafyffB1qK6-ahrTg) (I think The Meg and Mad Max are the brightest ones), and verify your findings? Cheers (I'll play with it during the weekend as well ...)

About NCL: I think it can be even useful for us with very "bright" titles where avgFALL is high, e.g. The Meg (see the sample above). What do you think?

Lastly, what do you think about this (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60322083)? (You can reply there to gain attention :D )

QBhd

11th December 2020, 12:54

I am sure that the S&M 10k Demo is about as tough a stress as you can get... it has a Max FALL of 1676. I will test the samples above tonight.

I will try and be a bit more clear on what I stated above. The horses in the snow is the prefect example of more DTN does nothing... it only takes about 10-20 DTN. Once you go higher, there is ZERO difference.

As for the negative effect, well I thought it would be clear, but the only negative effect I see with too much DTN is the image gets dimmer.

Okay so yeah... these two statements do conflict with one another.... Hmm how to explain this... Maybe it's the scenes that are not "homogenous" like the horses. Where there are dark parts with bright parts. We want to grab as much brightness without dimming the dark spots too much.

As for DPL, I am sticking with 1000 for now.

And I will reply to your post on AVS.

QB

QBhd

12th December 2020, 01:51

It's the Samurai vs QB! Battle to the death!! LOL

So I did a head to head with ALL 71 scenes with 150/100 vs 65/0 (DTN/NCL)... the final score: 7-36-28

Most of the 150/100 wins were so close as to be almost a tie, but there were a few highlights that would come to life, but really close. Almost all of the ties were either no difference, or I was flipping back and forth so many times I just called it a draw. And nearly all of the 65/0 wins were very clear cut.

At least to my eye and my tastes. There is no way to make this anything but subjective, I wish it wasn't the case :(

As for chros' sample files, the only time I saw 150/100 improve anything was with The Meg (clouds on the horizon were more detailed)... Almost all the other samples were nothing gained/lost. There were a few scenes in the mix where the dimming was very noticeable.

So I think I will keep the 65/0

QB

SamuriHL

12th December 2020, 01:56

I ran 65/0 for a while. I think I changed it when shadow recovery was being worked on but I just tried it again for a test I was doing for an audio issue and it seems fine. I can live with that.

And I just got told to start a new thread on AVS. Um, I've been part of that thread since damn near the beginning! I'm not some newbie posting "how do i configure this" questions for crying out loud.

VBB

12th December 2020, 02:02

You know how some of the PJ guys over there can get...

SamuriHL

12th December 2020, 02:05

He's more than welcome to search for my post history in that thread if he likes. LMAO

huhn

12th December 2020, 02:13

your screen is to small get used to it.

QBhd

12th December 2020, 02:15

LOL... Mine is even smaller!! :P 55" is plenty of awesomeness for me

QB

SamuriHL

12th December 2020, 02:22

We OLED users have helped them improve what they see on their low nit projectors quite a bit over the past couple years. What we ask for benefits them even when they can't see it. Cause their nit levels are too low. BAHAHAHA man I'm funny.

chros

12th December 2020, 09:55

And I just got told to start a new thread on AVS. Um, I've been part of that thread since damn near the beginning! I'm not some newbie posting "how do i configure this" questions for crying out loud.
Guys, just ignore (!) those comments, don't even reply and keep posting there (I don't even know who that guy is).

chros

12th December 2020, 12:23

your screen is to small get used to it.
:D That's a good one! But I think it's not even about the size but the prize! :D

I will try and be a bit more clear on what I stated above. The horses in the snow is the prefect example of more DTN does nothing... it only takes about 10-20 DTN. Once you go higher, there is ZERO difference.
Cheers, that was really interesting.

As for the negative effect, well I thought it would be clear, but the only negative effect I see with too much DTN is the image gets dimmer.
Oh, I see now what you meant :)

Where there are dark parts with bright parts. We want to grab as much brightness without dimming the dark spots too much.
Yes, makes sense.

As for DPL, I am sticking with 1000 for now.
OK, cheers.

So I did a head to head with ALL 71 scenes with 150/100 vs 65/0 (DTN/NCL)... the final score: 7-36-28

Most of the 150/100 wins were so close as to be almost a tie, but there were a few highlights that would come to life, but really close. Almost all of the ties were either no difference, or I was flipping back and forth so many times I just called it a draw. And nearly all of the 65/0 wins were very clear cut.
I think this test was useless :) since you already established before what different DTN values do. See below why:

As for chros' sample files, the only time I saw 150/100 improve anything was with The Meg (clouds on the horizon were more detailed)... Almost all the other samples were nothing gained/lost. There were a few scenes in the mix where the dimming was very noticeable.
Cheers. I think the big difference with The Meg is the used 100 NCL and not the 150 DTN.
Can you test this clip with different NCL values e.g. 0/100/300/500 (and maybe with couple of other DTN values, 65/100/150)?

So I think I will keep the 65/0
Sounds good to me, I'll play around with these as well, and thanks for your detailed tests!

QBhd

12th December 2020, 13:28

Actually The Meg, would have been improved with the 150 DTN, the scene is so bright that the NCL would have zero effect.

QB

chros

12th December 2020, 14:49

Hmmm, Interesting...

bkrieger

12th December 2020, 17:41

Hope that helps :)

Hello,
I would like to follow up on this. I had not been able to get this to work as the video had been choppy even on the lowest settings, but I was finally able to order an RTX 3080 from Newegg last night, and would like to know when it arrives what settings are recommended. Should I use NGU sharphigh on both Chroma Upscaling and image Upscaling?

Should the Nvidia Control panel color settings be set to 444?

Thanks

chros

12th December 2020, 19:58

As for DPL, I am sticking with 1000 for now.
Actually, about going really high: what about LG's ABL with HDR10?! :)

I went to rtings, to try to estimate the HDR peak with 0.01 window (based on their HDR peak measurements) for specular highlights, and I noticed (forgot about it) that they have a "HDR Real Scene Peak Brightness" value as well:
- B8 (https://www.rtings.com/tv/reviews/lg/b8-oled): 10% 693, 2% 722, Real-scene 548 (!)
- C8 (https://www.rtings.com/tv/reviews/lg/c8-oled): 10% 918, 2% 979, Real-scene 666 (!)
- E8 (https://www.rtings.com/tv/reviews/lg/e8-oled): 10% 868, 2% 929, Real-scene 685 (!)

When we measure 1% window the rest of the screen is completely black, so there's no ABL. But with a real scene ... :)
Obviously we don't know what scene they used for this (hopefully not The Meg :D) and how they measured the real peak but this is a valid argument.

So let's assume that 1000 DPL with C8 is fine with really dark scenes but with a normal scene it can only display 670 nits. In this case LG will clip everything above 670 nits :)

VBB

12th December 2020, 19:58

@bkrieger - Go crazy with that 3080 :D

I would set Nvidia to RGB full, 10-bit, 120Hz, CX to PC mode, madVR to 0-255, DCI-P3, gamma 2.4 (or whatever you use on the panel), automatic refresh rate switching to 2160p100, 2160p119, 2160p120, NGU sharp for chroma and luma as high as you can. Done! Let her rip ;)

chros

12th December 2020, 22:10

So, I tested those 12 samples (https://mega.nz/folder/n5dzgQbJ#A99MVeafyffB1qK6-ahrTg):
- madvr settings: 800 DPL, 65 DTN, 0 NCL (and the buggy dynamic clipping is disabled)
- testing methodology: using my remote (https://github.com/chros73/madvr-js-remote) to switch between passthrough (PT) and pixelshader (PS) while video is playing (!)

Results: :)
- as we know, LG's curve can make the image brighter with PT/PS if maxCLL is between 501 and 799 (?)
-- couple of titles are affected by this, e.g.: aba, thgm, gotg
- I haven't noticed any (!) difference with titles around 1000 maxCLL (LG's curve is the same with PS and PT)
- image gets a bit darker using PS with titles way higher than 1000: tm, mmfr, pr
-- PS has to compress a lot, resulting in a bit dimmer image oreverall

One last note about setting high, e.g. 1000 DPL in madvr using PS:
- don't forget that most of the options (even all of them?!) in madvr won't do anything (!!!) with frames / scenes that measured peak is below this 1000 DPL !!!
- that means that the majority of the scenes (not even titles!) you watch madvr basically only does passthrough!!!
-- how much of this percentage is with real content? I'd say at least 95% !!!!! :) That means on average 2 hours movie watching you only see madvr's PS effect for 6 minutes tops!!! :D

So, I go back to my used settings: 700 DPL, 55 DTN and the bloody dynamic clipping is disabled :)
- although the image is a bit brighter than the average passthrough but I don't mind :)

Next I'll try to take a look at NCL ...

bkrieger

12th December 2020, 22:50

Asmodian

13th December 2020, 01:00

I would set Nvidia to RGB full, 10-bit, 120Hz, CX to PC mode

You really do not have worse banding with RGB full compared to YCbCr limited in PC mode? Nothing I do prevents RGB full from having worse near black banding on my CX. It was the same on my C9, I don't think these LG OLEDs process full range video very well. :(

I want to use RGB full but the near black banding drives me crazy.

VBB

13th December 2020, 01:13

Well, I don't have a CX or C9, so I was just going by what I would do if I had one. If you say the banding is still worse, then I trust you 100%. I was under the impression that banding was a non-issue with the C9/CX in HDMI 2.1 PC mode, but apparently that's not the case then.

With my C7, I don't have any noticeable banding in SDR PC mode, and some noticeable banding in HDR PC mode, depending on the scene. This is with my current RGB full/TV low settings. Perhaps that does make a difference. Either way, I've switched back to normal HDMI mode, because I want one setting to rule them all! Unfortunately, I've been battling lip sync ever since :(

Asmodian

13th December 2020, 01:24

It isn't PC mode that is the problem, it is the TV using black level high that is the problem. I normally only test in PC mode but switching to HDMI mode to see if that would help does not seem to improve it. Normal banding is very good in PC mode on my CX, not sure using HDMI mode is any better.

LG also fixed the HDR in PC mode issues you have on the C7 for the CX but the banding I am complaining about is much less significant during normal scenes than that was. It is only noticeably worse with the behavior near black. Using madVR in limited and the TV in low would make all the difference, the issue is with the TV's behavior with black level set to high. :(

VBB

13th December 2020, 01:31

So, ideally you should do what I'm doing, but you also use your PC for other stuff.

Asmodian

13th December 2020, 03:11

bkrieger

13th December 2020, 07:58

Exactly. Your setup is ideal for madVR but I assume a lot of CX owners will also be playing games. :)

Edit:
It is the near black banding in games that I cannot get over too, though it is bad with compressed content in madVR as well.

Yes, I also play video games with it. So should I leave on PC mode and instead of RGB Full should I change the color settings in Nvidia control panel AMD set blacks in the tv settings to low?

chros

13th December 2020, 08:15

Exactly. Your setup is ideal for madVR but I assume a lot of CX owners will also be playing games. :)

Edit:
It is the near black banding in games that I cannot get over too, though it is bad with compressed content in madVR as well.

Why don't you use normal HDMI with game preset then for hdr? (there shouldn't be banding but get full chroma)

VBB

13th December 2020, 08:34

@bkrieger - What Asmodian is saying is that YCbCr 4:4:4 is the better option for you guys. PC mode itself is fine. Black level low, yes.

Asmodian

13th December 2020, 12:49

Why don't you use normal HDMI with game preset then for hdr? (there shouldn't be banding but get full chroma)

Hmm? Game mode does not give full chroma, only PC mode gives full chroma.

Also, PC mode is fine with HDR on the CX. The issue is that when using black level high with any mode I get significantly worse near black behavior.

quietvoid

13th December 2020, 21:41

chros

13th December 2020, 22:15

VBB

13th December 2020, 22:23

Not everyone ;) That is one issue the 7-series and earlier do not have.

quietvoid

13th December 2020, 22:29

I just tested black levels again the other day, I'll post here the results later.
But what do you see with the BlackPLUGE.U4K_DVp5 (https://bbs.io-tech.fi/attachments/blackpluge-test-patterns-zip.406872/) pattern?

I just see the right vertical bar, and barely the left one unless I increase the brightness.
Nothing on the left side at all other than some blobs when I get very close.

bkrieger

13th December 2020, 23:31

@bkrieger - What Asmodian is saying is that YCbCr 4:4:4 is the better option for you guys. PC mode itself is fine. Black level low, yes.

Ok, I will use that then.
On the PS5 it had been recommended to leave black levels on low or auto as well.

Thank You

SamuriHL

14th December 2020, 02:42

Are you calibrating DV at all?

VBB

14th December 2020, 03:06

Other than adjusting the high 2-point, no.

SamuriHL

14th December 2020, 03:19

So quietvoid isn't trying to upload a DV profile or mess with the LUTS? Cause that WILL cause black issues in DV on these OLEDS...at least some of them.

quietvoid

14th December 2020, 05:43

No.. I'm using the default settings, even reset the profile.

SamuriHL

14th December 2020, 14:25

Ok. Just ruling that out as an issue.

Sent from my SM-G975U using Tapatalk

chros

15th December 2020, 06:57

As for DPL, I am sticking with 1000 for now.
What settings did you use for these during your test?
- no compression limit (NCL) (mine 100) (I guess yours was 0)
- disable avgHighlights ceiling (DAC) (mine is enabled)

I guess you didn't enable DAC, it has a big impact with high FALL content: using 55 DTN + enabled DAC results in higher ADPL (dimmer image) than 65 DTN + disabled DAC.
And probably that's why you didn't see any effect when you raised DTN above a certain threshold (now I understand), but when you enable DAC then there won't be such threshold, it will continuously dim the image if you increase DTN.

Now the question is what we want with this DAC option? :)

Edit:
At least that's the case with S&M demo. But for example using the jw2 clip (@20 sec) it works the opposite way :) :
- FALL of the frame is only 16 (with sky strength 100, this setting also has an effect with this frame, starnge) and measured peak is 1464 (lights on the firetruck)
- 65+noDAC is dimmer than 55+DAC (the difference is not huge though, 8 ADTN)
- with this frame you can go up until 429 DTN (!) + noDAC here to reach the threshold!!!!

So I guess noDAC can help a lot with high FALL frames/scenes, not sure :)
And that's why most of the scenes of S&M didn't do much after the threshold because the algo already reached the avgHighlights ceiling due to the very high FALL.

And that's also a good example that S&M is not the go-to clip to check, it's not representative at all (more like an extreme case), it's just another content to check :)
If anything, that jw2 frame is, having dark scene with 1-2% window of really bright highlights.

QBhd

15th December 2020, 11:05

DAC was enabled for those DTN comparisons.

I have had a eureka moment recently in the way I think about "tone mapping" with pixel shaders.

With brightness on OLED being a variable thing I don't think we can attain the same goals that a projector user can or wants. So let's scrap that idea. Instead let's think about what we can do. One can make the argument that the pixel shaders are actually remastering the video frame by frame in real time. With that, one could then make the point that the pixel shaders are then a virtual "mastering display". So if we are in the remastering business, we don't want to do too much, but just enough to improve the content. Also, HDR comes in many different flavors and if we can remaster all these flavors to the same Vanilla that the OLED likes then we win.

QB

SamuriHL

15th December 2020, 14:39

My ub820 already does that.

Sent from my SM-G975U using Tapatalk

QBhd

15th December 2020, 15:34

Yes, but I doubt it is doing it frame by frame and in real time. I don't think it has the processing power to do such a thing. From what I hear, that player is awesome... for a player. :)

QB

chros

15th December 2020, 15:55

DAC was enabled for those DTN comparisons.
Thanks for the confirmation, it's good news! I thought for a moment that when you said "after a point there's no difference" that means the DAC was on.

So if we are in the remastering business, we don't want to do too much, but just enough to improve the content. Also, HDR comes in many different flavors and if we can remaster all these flavors to the same Vanilla that the OLED likes then we win.
I agree (we can't do much more about it :) ), the question is which settings are the best ... :) But I think our current settings not that bad either.

SamuriHL

15th December 2020, 16:17

Yes, but I doubt it is doing it frame by frame and in real time. I don't think it has the processing power to do such a thing. From what I hear, that player is awesome... for a player. :)

QB

It's nowhere near what madvr does, but you're talking about only enabling the bare minimum of madvr processing per frame. My point is, that if that's your goal the end result isn't going to look all that much different than what the UB820 already does. Essentially I think the biggest bang for our buck is the metadata change to force the C8 to behave with a curve that follows EOTF more strictly. The minimal processing on top of that isn't going to be super noticeable in a moving picture. In fact, I'd be super curious what your preferred settings look like compared to passthrough with metadata set to 1000/1000/400. Something I can actually do here.

QBhd

15th December 2020, 18:11

I'll freely admit that I am looking to do as little as possible. I don't think drastic changes to content is good or should even be a goal. The main thing I see is blown out whites (highs) in content that is very high. Getting more detail in those blown out parts is a good thing, as long as it doesn't hurt other parts or other scenes. Hence my conservative 65 DTN setting, it seems to do the job well, but it does struggle with certain scenes like the Meg... Even that scene is improved, just not quite enough.

The first thing I would do if I was SamuriHL. I would set the metadata to 1000/1000/333 and measure DPL=501. I would really like to know for sure that the massive jump in Luminance between 500 & 501 is actually the TV or if it is some threshold in madVR. It is a question I cannot prove. There is a definitive change in curves when changing the metadata with HCFR, but we cannot limit the Luminance with HCFR like we can with pixel shaders so the Luminance is the same, it just follows a very different curve. If you set DPL to 500 it only pushes 500 to the TV. Actually as I am writing this and "thinking aloud", I would need to do a few more measures to make a conclusion.... 501 with 500/500/167 meta as well would be another key one...

QB

p.s. added my current settings
http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/Settings.2.0.jpg

quietvoid

15th December 2020, 18:30

Shadow recovery at insane? What.
I thought your goal was to do as little as possible.
Although it only affects scenes over DPL I guess.

QBhd

15th December 2020, 18:46

Yeah, for the most part it will have little to no effect... I just want it to be aggressive when it is actually needed.

QB

SamuriHL

15th December 2020, 20:09

Actually, you may be surprised to find out that shadow recovery does a lot more than you may believe. Because when he's adjusting luminance it's having an impact in low black levels. I'm not really sure why that is and for us it really shouldn't be doing that. Anything that is belows 100 nits should be left alone and I THOUGHT there was a way to ensure that. We may want to look into that more cause I have shadow recovery set to high for similar reasons. It has been shown to make a difference. Look at some of the crazy SM scenes and you'll see what I mean. Anything that's very bright with shadows (duh lol) will show the issue quite well.

As for your measurement tests, I can probably do that but it may not be for a few days. I just don't have a lot of free time this week to mess with things. I see what you're trying to get at. And yes, I can set the DPL to whatever I want and then override the metadata with the Fury so I can definitely test that out at some point.

QBhd

15th December 2020, 20:28

While I was in the shower getting ready for work I realized that it should be an easy two measurements. 501 DPL with 500/500/167. And then 501 DPL with 501/501/167... This should prove it is the TV doing weird stuff.

QB

SamuriHL

15th December 2020, 21:32

I'll see what I can do. I might be able to slip that in tonight if I get a minute. No promises.

QBhd

15th December 2020, 21:35

It's also not that important for us. I am pretty sure we know it's the TV, but it's just a nagging doubt since I could not prove it

QB

SamuriHL

15th December 2020, 21:37

It's easy enough to test I just need a few minutes to get everything set up and test it. Shouldn't take more than 10 minutes once I get the meter set up.

QBhd

15th December 2020, 21:41

If you use HCFR you can post the L data and I can then plot it with Excel

QB

chros

15th December 2020, 22:21

Good teamwork, guys!

And if you haven't noticed it: the first (?) directshow HLG2HDR10 renderer (https://forum.doom9.org/showthread.php?p=1930936#post1930936) by mpcVR :) (along with HDR10 passthrough)
I think madvr got out of date :D

(I got stutter/tearing with Lav and bitstreaming, but works fine with the internal filters and audio decoding, I didn't have time to play around with it more).

SamuriHL

15th December 2020, 22:29

I don't have HCFR and don't have time to install/learn it right now. I use ColourSpace for everything these days. I do have Calmann too but I'd prefer not to use that.

SamuriHL

16th December 2020, 00:26

Wow that's some messed up junk!

DPL: 501

Full Field White 201 cd/m2 pattern (best I could do in a hurry to get it to drive through madvr)

500/500/167: 115
501/501/167: 132

Here's the thing...I was sitting on the SM advanced video menu when I made the metadata switch and I watched the TV "adjust" the brightness. It is 100% fact that it's the TV doing MESSED UP junk at 501.

SamuriHL

16th December 2020, 00:39

Now you guys have done bad things and got me thinking. That brightness jump is what we see when we enable dynamic tone mapping on the LG. Maybe it's throwing the damn thing into the 501-8xx curve??

QBhd

16th December 2020, 01:15

If I was to guess... I would say that at 501 the TV switches to a tone mapping that strives for the brightest possible image. 501 is well below what it is capable of... but still considered HDR. If you look at my first set of data where I measured the different DPL in the pixel shaders ... Everything above 501 was pushed to 700+

QB

SamuriHL

16th December 2020, 01:29

Well I have zero intention of allowing the metadata to ever fall back into that range again. That's absolutely stupid. I still maintain we want an actual DPL for madvr to tonemap to with some allowance for specular highlights to go above that up to the metadata we set of 1000/1000/400. I know you all say that just setting DPL to 1000 and calling it a day is good enough but I'm not as sure about that as you guys are. Because I maintain that some of those settings we say don't impact us actually do on higher nit titles (mad max, meg, whatever). There was talk at one point between soulnight and madshi about allowing madvr to tonemap specular highlights separately from the rest of the frame. THAT is what I want. Exactly that.

EDIT:

Consider this example of a frame. A very bright sky of say 1500 nits with the sun as a light source of say 4000 nits. I want the sky tonemapped to the real display nits and the sun to be able to tonemap slightly above that. I'm picky like that. :)

quietvoid

16th December 2020, 01:50

Isn't that how the UB9000 works? At least it looked that way in Vincent's video, where only the highlights were mapped down.

SamuriHL

16th December 2020, 01:54

I don't fully know what the Panny players are doing. I can see that they're setting the metadata to 1000/1000/400 (more or less) for OLED devices. Then they have those dynamic tone mapping tweaked settings that I tend to stay away from because they overly brighten things artificially. It's entirely possible though.

QBhd

16th December 2020, 02:10

The thing is, I don't think you can actually know how bright any pixel can get in any frame in any scene. Even a 1% window is huge. Brightness on OLED is far too variable. I think it is a fair comprise to remaster to 1000/1000/333. Even then, do we know that a flashlight in a dark room can't go above 1000?... I don't think it is a knowable value

QB

SamuriHL

16th December 2020, 02:18

Well we have no choice but to compromise right now. We may never get to where I want to be. Or by the time we do, I may finally get to buy the screen I really want. I want something with the picture quality of OLED but 4000 nit brightness. Then all this tonemapping stuff can go away. Sure, there's still 6000 nits to the max 10000 but really if we can get to 4000 the range is so high I'd pretty much be ok with on device tonemapping at that point for anything over 4000.

Asmodian

16th December 2020, 02:39

Over 4000 nits is seriously diminishing returns, I assume given that I obviously haven't seen a 10000 nits OLED. :D

I do think 2000 nits is where brighter starts to really not make sense, at least in a dark room with a self emissive display tech. At some point the dynamic range of human vision really does come into play and the engineering and power simply is not worth it.

In a fully dark room I am not sure I want my display to even hit 2000 nits full screen white, it would be very realistic in that a midday beach scene immediately after a night scene would make my eyes water and I wouldn't be able to see for a bit, but is that actually a better viewing experiance? I have had friends ask me to turn down the brightness of my OLED in HDR mode because it was too bright, and at 100% it still doesn't get much over 700 nits with a 10% window.

SamuriHL

16th December 2020, 02:50

I want the tech that can do 4000 nits. At that point, no tone mapping will realistically be needed. I hear what you're saying. There isn't a lot of content out there that's above 2000 anyway. So realistically, we're going to mostly be in that range. I just want the next display I buy to not be limited to tone mapping solutions.

chros

18th December 2020, 21:04

I'm going out of context but have any of you tested their black levels with Dolby Vision?
On my C8, there's black crush whenever the TV is in Dolby Vision mode, which is really annoying.
My Firestick 4K also has the same behavior, though I have to reduce brightness to 49 to avoid raising the black floor as well.
The TV internal apps are a bit better, but there's still an adjustment necessary.

The same pattern in HDR10 is perfect.

If anyone wants to try, pattern from testing-video samples: https://mega.nz/file/1V9mURbS#DPeqP1QlQtOCaUks-9lFxyo2urR43sKe0ISXMlPuNco

Edit: Well I found more info by @chros, apparently everyone suffers.

I just tested black levels again the other day, I'll post here the results later.
But what do you see with the BlackPLUGE.U4K_DVp5 (https://bbs.io-tech.fi/attachments/blackpluge-test-patterns-zip.406872/) pattern?

I just see the right vertical bar, and barely the left one unless I increase the brightness.
Nothing on the left side at all other than some blobs when I get very close.
Same here with 50 brightness, and probably the black is elevated. If I set brightness to 49 then the whole frame is pure black :)

I made the following test because someone stated on another forum that since HLG HDR and HDR10 shares the same preset mostly on all TVs (not just LG), one of them is not perfect. And as it turned out he is right.

Summary (at first), it seems that PC mode cuts black with every format (where it's applicable):
- SDR: PC mode cuts a little bit, doesn't matter whether limited or full the signal is, but it's not that visible
- DV: it cuts black a bit (adjusting brightness doesn't help)
- HDR10: PC mode cuts it, doesn't matter whether limited or full the signal is, but not that big
- HLG HDR: cuts a lot :) (works fine otherwise, LG reports HLG HDR)

Near black test on B8 with the good old 4.10.25 firmware, with these settings (https://www.avforums.com/threads/lg-b8-oled-owners-and-discussion-thread.2188352/post-27944812) (brightness 50):

Used settings:
- madvr + hdmi + *PC mode Full: TV black level High, nvidial RGB Full 12bit output, 3dlut is disabled
- Oppo + hdmi + *PC mode Limited: TV black level Low, Oppo RGB Limited 10bit output

SDR (tested only with madvr + hdmi):
- avshd-bt709 11-Black-Clipping:
- - madvr + hdmi + non-PC mode Full: from 1 meter 17 and 18 visible
- - madvr + hdmi + PC mode Full: from 1 meter 18 visible, 17 only visible from closer
- BlackPLUGE-U4K:
- - madvr + hdmi + non-PC mode Full: all of them are visible (the 2 vertical bars and 2 left rectangles (top and bottom) are a bit lighter, plus the 3. rectangle in the middle is also faintly visible)
- - madvr + hdmi + PC mode Full: 2 vertical bars and 2 left rectangles (top and bottom) are visible

HDR10:
- Mehanik 01. black-level-v1:
- - madvr + hdmi + non-PC mode Full: all of them are visible (66 too!)
- - madvr + hdmi + PC mode Full: 76 visible, only the frame (!) of 72, below 72 nothing is visible
- - Oppo + hdmi + non-PC mode Limited: all of them are visible (66 too!)
- - Oppo + hdmi + PC mode Limited: 76 visible, only the frame (!) of 72, below 72 nothing is visible
- - Plex + internal LG player Limited: all of them are visible (66 too!)
- BlackPLUGE-U4K_HDR10:
- - madvr + hdmi + non-PC mode Full: all of them are visible (2 vertical bars and the 3 rectangles on the left)
- - madvr + hdmi + PC mode Full: only the rightest 1 vertical bar is visible
- - Oppo + hdmi + non-PC mode Limited: all of them are visible (2 vertical bars and the 3 rectangles on the left)
- - Oppo + hdmi + PC mode Limited: only the rightest 1 vertical bar is visible
- - Plex + internal LG player Limited: all of them are visible (2 vertical bars and the 3 rectangles on the left)

DV BlackPLUGE.U4K_DVp5:
- Oppo + hdmi + non-PC mode Limited: only 2 vertical bars are visible
- Plex + internal LG player Limited: didn't recognise the file :)

HLG BlackPLUGE-U4K_HLG10:
- Oppo + hdmi + non-PC mode Limited: nothing is visible
- Oppo + hdmi + PC mode Limited: nothing is visible
- Plex + internal LG player Limited: nothing is visible

quietvoid

19th December 2020, 01:11

Thanks. Didn't expect HLG to be the worst, haven't noticed anything when I played some.
Apparently Dolby escalated the Dolby Vision levels issue: https://www.avsforum.com/threads/dolby-vision-is-broken-2020-lg-oled.3175856/#post-60324371

RXP

19th December 2020, 06:41

chros

19th December 2020, 17:16

Apparently Dolby escalated the Dolby Vision levels issue: https://www.avsforum.com/threads/dolby-vision-is-broken-2020-lg-oled.3175856/#post-60324371
Yes, I get the same thing with DV as D-Nice explained it on that sublink:
"0% stimuli is elevated (glows). Reducing the brightness to 49 corrects the glow. However, brightness 49 clips 0.5 - 2.3% stimuli)"

The interesting part is that why only the vertical bar(s) are visible and not the rectangles on the left (happens in PC-mode with HDR10 as well) when they supposed to have the same values?
It's like it treats the middle of the screen differently :)

Now you guys have done bad things and got me thinking. That brightness jump is what we see when we enable dynamic tone mapping on the LG. Maybe it's throwing the damn thing into the 501-8xx curve??
I don't think so that's the reason, they can easily achieve this another ways.
But the question is whether other years models work like this or not? (2017/2019/2020)

bkrieger

20th December 2020, 02:07

@bkrieger - What Asmodian is saying is that YCbCr 4:4:4 is the better option for you guys. PC mode itself is fine. Black level low, yes.

So I received the RTX 3080 and now everything is working.

A couple questions- Would there be much of a difference if I set Chroma Upscaling and Image Upscaling to NGU very high vs NGU High

Under HDR is it still ok to leave Passthrough to display or do I need to mess with the Tonemapping settings? Those seem confusing

Thanks

VBB

20th December 2020, 05:44

Set chroma and luma as high as your card can handle, but you probably won't notice a difference between high and very high. With the new card you can definitely do tone-mapping, but it is much more involved than simple passthrough. If you want more info, see here: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/

quietvoid

20th December 2020, 16:20

quietvoid

20th December 2020, 17:56

Tested quickly with the recent dicussions for madVR, seems that using 700 nits makes the image dimmer (at least for the S&M demo) compared to 1000 nits. Can't see anything wrong with 1000.
Got scared when there were flickering red pixels in shadows, replugged HDMI and it seems to be fixed.

I still have to play with saturation to see if the first shot mountain peak can be improved, it's very saturated using SamuriHL's settings.

chros

20th December 2020, 18:00

On my set there's no discernible difference starting at 72, and then lower bars blend in with the non black background, at brightness 50.
When lowering to 49 I get absolute black but bars below 83 disappear. :)

Seems to be perfectly aligned with the 0 - 2.3% clipping, at least!
Now I don't know which compromise to pick.
:) Same here from the Oppo. Well, HDR10 is slightly better in PC mode with 76 visible, so dunno. Probably I'll leave it as it is.

SamuriHL

20th December 2020, 18:02

Of course it is. That's what madshi has to fix and why I keep saying that even though we have high nits to play with compared to PJ owners we're still impacted. You can fix it by setting desat 1+2 + 50% dumb, but, I don't really recommend it.

quietvoid

20th December 2020, 18:53

Increasing desat doesn't help for that scene, I do remember it looking different but oh well.
Happy with 1000 nits target, 150 DTN. Clipping brings a bit more brightness but meh.

QBhd

20th December 2020, 23:51

You should try my settings... 150 DTN is too much compression.

https://forum.doom9.org/showthread.php?p=1930940#post1930940

QB

bkrieger

21st December 2020, 00:36

VBB

21st December 2020, 00:56

I would leave whatever you can as "let madVR decide", but feel free to play around with it.

bkrieger

21st December 2020, 01:10

I would leave whatever you can as "let madVR decide", but feel free to play around with it.

OK, I’ll do that.

Thank You

quietvoid

21st December 2020, 01:57

You should try my settings... 150 DTN is too much compression.

https://forum.doom9.org/showthread.php?p=1930940#post1930940

QB

150 is not too much at all.
While some scenes of the demo benefit from some higher brightness with setting it lower, others lose contrast and highlight detail.
The one I noticed was the waves hitting the rocks, there's detail lost with DTN at 50.

SamuriHL

21st December 2020, 03:14

Yea, DTN can be a pain in the ass to dial in properly. The last time I went through this exercise, which admittedly was quite a while back, I settled on 150 as it generally looked the best in most content. I've run with 75 before, as well and that handles quite a bit of content fairly well. It can be tricky.

EDIT:

Ya know....I never tried matching DTN to max nit level of the video I was playing. I wonder if there's a pattern there. E.G. the SM is 10k nits which is super freaking high and all over the place. But in real world content, I wonder if having a DTN for 4000 nit content and another for 1000 nit content makes sense. I suspect we could do that with profiles. Never dawned on me to even consider that.

QBhd

21st December 2020, 06:31

The S&M Demo does come in all kinds of flavors... there is a 600 MDL, 1000, 2000, 4000 , and 10000... Plus various colour spaces too... and a few other differences. There is 13 different versions.

But here is the thing... If it is not needed, more DTN will do nothing for us. So the lower nit sources will not need as much. I went through ALL 71 scenes, the handful of scenes where a higher DTN than 65 helped is not worth the more than 1/2 of the scenes where it got dimmer.

QB

quietvoid

21st December 2020, 15:46

I mean profiles would be great by MaxCLL, but it would probably be a bit painful on maintenance and keeping them updated.
I'd probably do 75/1000, 100/2000, 125/4000, 150/10000 or something.

chros

22nd December 2020, 12:37

150 is not too much at all.
While some scenes of the demo benefit from some higher brightness with setting it lower, others lose contrast and highlight detail.
The one I noticed was the waves hitting the rocks, there's detail lost with DTN at 50.
What about 65/75?
If it is not needed, more DTN will do nothing for us. So the lower nit sources will not need as much. I went through ALL 71 scenes, the handful of scenes where a higher DTN than 65 helped is not worth the more than 1/2 of the scenes where it got dimmer.
There's a big difference between 65/75 vs 150! :) (Note, I haven't gone through it so detailed as QBhd did.)

But here are couple of things that I don't understand:
1. DTN should be only (!) responsible of taking (high) FALL into account!
- it doesn't do anything with frames that their peak is below DPL: so if you use 1000 DPL then it doesn't do anything with frame peak below 1000 nits
- when DTN is disabled, it should not affect content with "normal" FALL, madvr's other algo should take care of it
2. high fall has nothing to do with maxCLL
- although you can argue that low maxCLL (~700) titles definitely won't get high FALL
- take a look at this list (https://docs.google.com/spreadsheets/d/1-t8_mF-jxTbArQIvNHm7XQUcMFUTvJ5O87bLFS1lCm0/edit#gid=1677456743), order by avgFall
- SM 10000 HDR10 has these values: MaxFALL: 1667, AvgFALL: 311, AvgFMLL: 3576 nits
3. if you still (!) feel the need of profiles based on whatever attribute, then it only means one thing:
- the algo that madvr uses for DTN with HDR passthrough is wrong!
- in this case you have to ask madshi to modify it according to your needs

QBhd

22nd December 2020, 23:52

I see zero need for profiles with regards to DTN

QB

SamuriHL

23rd December 2020, 03:16

Until we take a look at if they can truly help or not, I'm not ruling them out. If you have no need for them, that's great and it simplifies your configuration.

SamuriHL

26th December 2020, 04:30

Build 115 is interesting. That new option to not apply peak nit to DTN does some crazy stuff for me at least. I can't get the brightness to come back down to something reasonable no matter what I set the DTN to after enabling that option. I had to disable it. After playing around and a fair bit of testing tonight, I'm back to using DTN 150. I'm also using desat i + 75% dumb as it gives just a slight edge of saturation over desat i 1+2 + 50% dumb. desat i 1+2 + 50% dumb certainly seems correct technically speaking from what I can see. But for now I'm running desat i + 75% dumb just to play around with it. Hopefully someone can tell me what the hell the do not apply peak is supposed to do. I don't know if it's something we should care about or not but so far for me, not is the answer.

Magik Mark

26th December 2020, 05:57

Guys Can you share ur new setting under b115. Thanks a lot

Sent from my SM-G988B using Tapatalk

quietvoid

26th December 2020, 16:42

I've played with the new DTN option a bit but at a certain point it stops changing the still image (same as just changing DTN sometimes), gotta test more :)
I think we're supposed to be comparing the target tonemap nits on the OSD.
Oh well, looks like Neo-XP is testing too :)

Guys Can you share ur new setting under b115. Thanks a lot
There's not much that changed, probably just have to reset the saturation option (mine defaulted to y' after changing to 115).

SamuriHL

26th December 2020, 17:29

As mentioned, I'm using Neo-XP's desat i + 75% dumb and I like it so far. it gives just a tad bit more saturation than desat i + apply 2 + 50% dumb which is good for us IMO. YMMV of course. Also I had to up the DTN to 150 even with this new do not apply peak nits setting disabled. I need to play with that some more. I don't think that should have changed without the option enabled. I can tell you with that option enabled nothing I did in the DTN setting made the brightness even remotely ok. It was absurdly bright for me.

aron7awol

27th December 2020, 05:37

I don't actually see any saturation issues on our high-nit displays. I prefer desat i 1 only, no dumb. Is there some specific content that you guys have seen saturation issues with, even with our ~1000 nit DPLs?

I did some extensive testing with BvS and I definitely preferred NOT checking "disable avghighlights ceiling". I find the ceiling is a good thing, as it prevents DTN from kicking in too strongly. I really liked this combined with 75 DTN.

DPL 1000
DTN 75
unchecked "disable avshighlights ceiling"
desat i 1 only
Pretty much everything else disabled.

chros

27th December 2020, 07:29

Hey Aron, welcome here!
What about the new killer feature "don't add DPL" ? :) That's the biggest change for us since DTN was invented!

SamuriHL

27th December 2020, 15:42

Are you able to use it without overblowing the brightness?

Sent from my SM-G975U using Tapatalk

aron7awol

27th December 2020, 17:06

I haven't tried the "don't add DPL" feature yet. It sounds like maybe there is a bug and it isn't working right, but it's not really a big deal. It mostly just makes it a bit more consistent to dial in DTN across different DPLs, and if you guys settled on using a particular DTN before, you'll probably have to experiment to find your new DTN setting with that box checked. But in my testing I really liked the avghighlights ceiling kicking in when it did. It kind of makes DTN a bit less important if DTN is set high enough to hit that ceiling in a lot of cases, but in scenes that don't hit the ceiling there are still noticeable differences between different DTN settings, especially lower ones like 50 vs 75.

Trying to figure out what is highlights and putting a ceiling on the TM target always made sense to me in principle, especially on high-nit displays. It essentially avoids unnecessarily dimming the overall picture for highlights only, but still will dim the picture if the whole scene is really bright and it's not just highlights, and in that scenario we need to dim the picture to fit it all in with good detail. This results in a much more consistent experience as far as brightness, especially on a high-nit display that can really handle 95% of the scenes out there without having to dim the overall picture, and only when we have scenes that the whole scene is pushing the limits do we need to dim everything (which ABL would likely do anyway).

chros

27th December 2020, 21:31

I agree, and that's why the new feature that you and neo-xp asked for does the same thing.
Check that box and set 60 DTN (I used 55 without it previously) and try out different scenes/movies/tv shows. Don't forget that with high DPL (e.g. 1000) DTN is not used at all if the measured peak is below the set DTN.
The result is just amazing! No more unnecessary dimming with low FALL titles!
My opinion is that this is how it should have worked from the beginning! And that means we watched the wrong images in the last 1.5 years! :)

And here's a riddle for you guys: try to figure out what formula the DTN "algo" uses :D
But seriously, ask madshi about it, because it's insane that we have to fool around something to try to figure out how it should work...

SamuriHL

27th December 2020, 22:33

If any of you have the SM disc, here's what I want you to do.....set your favorite settings, whatever you want. Now, play the 1000 nit version of the demo file and go to 6:04. You'll see a little croc on some rocks. To the upper left there's some yellow and black mixed in with the rocks. Now check it out in the 10000 nit version. Do they look the same?

aron7awol

27th December 2020, 22:36

Well, I just felt that eliminating DPL from the calculation made for more consistent DTN results across different DPLs. I made a spreadsheet when we were discussing it so I could compare the results with both formulas. Even with it eliminated, I still feel the avgHL ceiling is beneficial, as it adds some additional intelligence as far as when to dim the whole image.

I'm glad to hear it seems to be working for you though, maybe there is no bug after all. :)

madshi alluded to there being a slight tweak to the formula, but didn't share it, so I got the feeling he didn't want to for whatever reason.

After my recent testing and posts in the AVS thread, I'm also thinking don't desat might actually be a reasonable option for us TMing to 1000 nits.

chros

27th December 2020, 23:56

We don't know whether there's a bug or not because we don't know how it's supposed to work :)
Maybe asking him via PM is more fruitful.
I think desat2 is needed even for us, again, these options don't do much when frame peak is below DPL, but when it's above then we have the same issue as the projector guys.

avgHL ceiling is interesting, never played with that one. Here's a bit historical overview:
- it was created by soulnight along with the FALL algo (DTN) for pj users
- madshi modified the original DTN algo and stated that this option is maybe not needed anymore
I think this option works the most when there's high FALL so "higlights" are not even. But when we got only couple of small light source it doesn't do much.
But again, we don't know how he defined "highlights"... :)

@SamuriHL, I don't have the 1000 version, if you can upload it then I can check.

aron7awol

28th December 2020, 00:05

Well I think we know how it's supposed to work close enough to know if there's a bug or not. The difference between what we calculate based on the "known" DTN formula and what madVR reports has been very close.

I'm not referring to frame peaks below DPL, of course saturation is irrelevant in those cases as it doesn't tone map. But the scenes I tested in BvS and MMFR all had peaks well above DPL, and less desaturation looked better. Then I tested MMFR with the LG TM and madVR with the different desat options and don't desat was closest to LG's approach, and it not only looked better but also matches the non-TM version best. Now maybe a small amount of saturation (less than 100% desat 1) would be the best balance, I'll have to test more.

I believe highlights are defined as everything >100 nits for the avgHL calcs.

chros

28th December 2020, 01:09

"highlights are defined as everything >100 nits for the avgHL calcs"

I hope not (don't think so), because when the whole content is shifted (like The Meg, SM 10000 demo, etc) it wouldn't work. That's what I meant how he defined it.
But maybe that option is useful indeed, I also have to play with it and see, but with normal content and not with demos :)

About desat2, you saw the discussion over there, it's not that simple. I think James is right, fire should be yellow not the nonTM version!
And I wonder whether your CX acts the same with TM as our 2018 models :)

aron7awol

28th December 2020, 04:18

I did some testing with The Meg, and I actually preferred it with the ceiling in most shots. Obviously this stuff comes down to personal preference, but I actually am really liking when it is kicking in, it makes the image look more realistic to me, while without the ceiling it dims too much and no longer looks like an outdoor scene. Or as another example inside the elevator, with the ceiling there is a really good balance between the brightness of the light and the brightness of the walls and people, it looks realistic, but without the ceiling, it dims and the light is still really bright but everything else is too dim relative to it. FWIW, I did a bunch of A/B comparisons for my wife from The Meg, who has no idea what I was even doing other than just asking her which she preferred, and she chose the ceiling every time. The same happened earlier today when I did the same with some stuff from BvS. So all in all, I'm liking it and I actually think it was tunable rather than just 2xavgHL it could be really great and useful even for people that might prefer a bit dimmer than me but could still take advantage of the ceiling. I have similar feelings as you just described with the DTN algo tweak: no more unnecessary dimming!

Yeah, the desat/hue stuff is a whole other can of worms, and I actually think that the choices for TMing to 1000 nits or 100 nits maybe should be drastically different. It's almost as if the hue shift helps offset the flatness of the heavy compression when going down to 100 nits, which does seem potentially desirable. But for us and TMing to 1000 nits, I actually preferred the no or very little desat approach, and I think having ~1000 nits to throw at it is the difference there. Because while everything looks pretty flat in my terrible photos in that thread, having that fire with ~1000 nits it looks totally not flat and lively and realistic. FWIW, my Sony does shift some of the fire to yellow even with its 1000 nits, and it doesn't look as good to me when comparing side by side. Since I got the CX I've had it and the Sony side by side on dual outputs from my AVR so I've been able to compare a ton of content that way, which has been great.

We just watched Ghostbusters (2016) with the avgHL ceiling and don't desat and the results were phenomenal. But obviously without a frame of reference that's not worth much, so after the movie was over I went back to a whole bunch of the brightest frames and compared with/without the ceiling, and separately with/without desat. I once again preferred the ceiling and don't desat, and this has become my normal setting.

SamuriHL

28th December 2020, 05:49

I can't upload that demo video unfortunately. But that disc is worth having for this very reason. Looking at it at 1000 gives us some sense of what it should look like when tone mapped from 10000 -> 1000. Right now for testing purposes I have all the enhancements off, ceiling enabled (I agree, that's an interesting thing to play with....I've always had it off), DTN 65 and I forget what my desat is set to. most likely desat i + apply 2 at the moment. That scene at 6:04 looks very wrong to me. I can't tell but it almost looks like it's shifting the white to almost pink in some parts and it's WAY overexposed and loses the detail. Now I turned highlight recovery, shadow recovery, etc all off right now but even with some of those on I could NOT get madvr to tone map that to look correct. Which is why I want to see if anyone else can and what settings they use. This is very much a torture test I know. But it's indicative of the problems we face when even doing 4000 to 1000 tone mapping.

aron7awol

28th December 2020, 06:18

I'll take another look at the S&M disc with my new settings and report back.

On the highlights ceiling, even in a case of a scene from The Meg with really high FALL, avgHL will always be equal or higher than FALL since it is basically FALL but without the <100 nit pixels. And so the ceiling ends up as double that which is still quite high on those high FALL scenes. Now maybe 2x isn't quite enough, I'd love to have to configurable.

chros

28th December 2020, 13:53

I can't upload that demo video unfortunately. ... Looking at it at 1000 gives us some sense of what it should look like when tone mapped from 10000 -> 1000. Right now for testing purposes I have all the enhancements off, ceiling enabled (I agree, that's an interesting thing to play with....I've always had it off), DTN 65 and I forget what my desat is set to. most likely desat i + apply 2 at the moment. That scene at 6:04 looks very wrong to me. I can't tell but it almost looks like it's shifting the white to almost pink in some parts and it's WAY overexposed and loses the detail. Now I turned highlight recovery, shadow recovery, etc all off right now but even with some of those on I could NOT get madvr to tone map that to look correct. Which is why I want to see if anyone else can and what settings they use. This is very much a torture test I know. But it's indicative of the problems we face when even doing 4000 to 1000 tone mapping.
This is my b115 settings (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-438#post-59228954): I also disabled contrast recovery as well, sometimes it ruins it :)
So basically almost all the options are disabled due to bugs in them or just don't work at all. :)

And I disagree with the validity of the 10000 vs 1000 test as I said above, due to this (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60365341).
And what madshi said when he was back 1 month ago, what do we want from madvr to do: keep the nonTM image or fix the "broken" master? :) In this case (mad max explosion(s)) the latter :)

On the highlights ceiling, even in a case of a scene from The Meg with really high FALL, avgHL will always be equal or higher than FALL since it is basically FALL but without the <100 nit pixels. And so the ceiling ends up as double that which is still quite high on those high FALL scenes. Now maybe 2x isn't quite enough, I'd love to have to configurable.
You convinced me, I also disabled the "disable avg HL" (as you can see above), this is when I played with it (https://forum.doom9.org/showpost.php?p=1930893&postcount=1208) a bit.

I made a spreadsheet when we were discussing it so I could compare the results with both formulas. Even with it eliminated, I still feel the avgHL ceiling is beneficial
I'd be interested in it if you can share it (e.g. via google spreadsheet), here's my simple test result with "don't add peak nits":

FALL DTN to reach >700 DPL
15 -
24.27 -
24.58 500
24.75 496
27 457
33 369
42 295
49 249
67 184
76 162
100 123
130 95
161 77
207 60
245 51
286 43
353 35
449 28
730 17
842 15
900 14

FALL DTN to reach >800 DPL
27.48 -
27.7 497
28 491
220 63
231.64 60
237 59

SamuriHL

28th December 2020, 14:40

The difference in this case on the 10000 vs 1000 is that not only do we have it in hdr we also have it in dv. And it's not broken in dv. Madvr is doing something whacked here.

Sent from my SM-G975U using Tapatalk

chros

28th December 2020, 16:12

Hmm, doest it change with the differernt desat options?

SamuriHL

28th December 2020, 16:38

I need to test all of them but the ones I tried...no, not really. None of them fixed the issue.

SamuriHL

28th December 2020, 16:49

Ok, this is definitely a problem. The issue goes away when I enable compromise ON. WTF? This is messed up. Whatever madvr is doing in the hue shift with compromise off is screwing this up. I even tried desat i and nothing else enabled and it's still wrong.

chros

28th December 2020, 16:55

Cool, tell madshi, post pictures etc,...

And if you guys are womdering why I tell You guys to do it:
none of my 2 recent bugs were replied (by anyone, but you, when I asked You guys), let alone fixed. Hence I stopped participating there.
If you still be interested in those and other possible future observation of. mine, I can post them here, but You have to force to be fixed :)
I'm tired of this ....

SamuriHL

28th December 2020, 17:21

I hear you. There's a lot of...."stuff" going on lately that isn't awesome. I finally fixed the ability to grab screen shots so I will go post them.

chros

28th December 2020, 17:41

Cheers, we can still work together here for the greater good if that's what we want... :)

SamuriHL

28th December 2020, 18:10

I think we are going to have to if we want to drive these OLED panels to their best selves. LOL We're constantly learning and they're constantly changing things that require us to retest our previous assumptions.

aron7awol

29th December 2020, 20:56

SamuriHL

29th December 2020, 21:43

Yea I saw some people playing with that. I've not had a chance to test max yet. Will give that a go tonight.

chros

29th December 2020, 22:36

I just posted in the AVS thread, but after some additional testing, I'm finding max to be SO much better than sep. It fixed all the issues I was having with too much desat, even with just desat i 1. I think that combo is perfect for us with OLEDs.

Cheers, but take a look at the mmfr sample from 0:15-0:30 : top left, big reflection on the edge of the car.
With max+desat-i+nodumb (no desat2): those 2 big patches become blue! Nothing helps it.
This is way more annoying than some color shift.

Also at 2:15 the sundown on the left, becomes orange (but we know that it's probably the source.)

NeoXp reported issues with max couple of weeks ago and if I have to bet, nothing happened to it.

aron7awol

29th December 2020, 23:07

I don't have the sample clips. Do you know the timestamps from the movie?

FWIW, linear 54 was almost identical to max. Does that have the same issues?

chros

29th December 2020, 23:53

How come you don't have the samples?! :D

It's in the beginning, 00:01:20.
And yes, 54 is similar to max.

aron7awol

30th December 2020, 00:04

How come you don't have the samples?! :D

It's in the beginning, 00:01:20.
And yes, 54 is similar to max.
I didn't see where they were posted/hosted.

aron7awol

30th December 2020, 00:17

Ok, I see the reflection in the car, and I see that it is definitely blue with max, but it is also definitely very blue in the source! This is what I found in my other testing: max is very true to the source. And FWIW, it is sep that can create artifacts in the green spear for me while max looks true to the source there as well.

SamuriHL

30th December 2020, 02:07

My disc just came so I am watching it now with these settings. max, desat i, highlight 1000, DTN 65. Everything else shut off. I've only ever seen this in DV streaming so...

SamuriHL

30th December 2020, 03:34

I forgot what a messed up grade this is (I've not watched it in a while and as mentioned it was DV streaming). These settings seem quite faithful to the grade. You aren't going to fix the source with these settings, but IMO it's not for us to fix. I'm rather impressed.

aron7awol

30th December 2020, 04:30

aron7awol

30th December 2020, 04:32

SamuriHL

30th December 2020, 05:28

So I just tested this scene. This is one where the 2xavgHL ceiling is not high enough, and it really needs to be 2.5x or 3x. Then it looks quite good to me and very similar to the 1000 nit grade.

That would explain it. Good. Maybe madshi will consider letting us set a scale for the avgHL ceiling. 3x would probably cover everything.

SamuriHL

30th December 2020, 05:31

Yeah, MMFR really has quite a jacked up grade! But I agree, if we start messing with our settings to make it look more "normal" we'll just end up screwing with everything else. So I'm with you, it's not for us to fix, let it be what it is.

I realize this is a huge philosophical discussion but I fall in the madshi camp on this one. We want to reproduce what's on the disc, not what "looks good". If people want to chase that elusive "fix the source" viewing, have at it but allow us who want to remain faithful to what's on the disc to do so. If the person who grades the movie has no f***ing clue what they're doing or using s****y equipment to grade on, so be it. LOL I want accuracy.

chros

30th December 2020, 11:33

:D OK, so we all agree that mmfr is screwed up. There's indeed a small blue patch inside of those reflections at 10000 DPL.
But this is interesting, it means we can use the following to correct bad master (e.g. manually via profile): sep+desat-i+nodumb+desat2

OK, I'll switch to max+desat-i+nodumb , and will create a profile for the above.
@Aron, does the highlight setting (0-1000) matter in this case? (And do you know about profiles in madvr?)

And I linked the samples here (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60322083) for you.

aron7awol

30th December 2020, 18:17

Yeah, I don't see a problem with a profile for a specific really strange of bad grade like MMFR. sep instead of max, maybe fire tweak, whatever else makes sense. After all, what looks best is the #1 priority for me always. But I just don't want to adjust my daily driver settings because of a bad master, which I'm sure you guys feel the same way on.

I didn't notice any difference from the highlight setting in some quick testing, but I didn't do too much so I'm sure there are some instances it has an effect? If you guys know of any or find any, please share.

Yes, I use profiles all the time to switch back and forth for comparisons. That's why when I saw the guys on AVS trying to compare the desat stuff and using side-by-side different displays and all that, I was wondering why they wouldn't use HDR output and just toggle between profiles via hotkey. It's by far my favorite way to compare all of this stuff.

Thanks for the link to the samples!

SamuriHL

30th December 2020, 19:31

I see that madshi agreed to your avgHL multiplier option. That should be fun to play with. I had the highlight set to 1000 for MMFR last night and didn't seem to make a huge difference, but, I was watching the movie rather than pixel peeking so it's possible it had an impact. But if so it wasn't glaringly bad.

It's a lofty goal to try to create profiles to fix grading/source issues to make them look better. The only time I'd really want to go there is if the grade is so egregiously bad that it completely pulls you out of the movie. MMFR, as whacked as the grade is, seems fairly intentional. Atomic Blonde is another crazy grade which is very much intentional. Going around "fixing" these....I don't know, just not my thing I guess.

chros

30th December 2020, 22:51

@Aron "What is the actual formula?"
:D Thanks, that's why I said couple of days ago that I pay a drink if someone can figure this out :D
Also said, that when he implemented soulnight's DTN formula, he changed it, only god and he knows to what :)

And that's why I chose 60 DTN, if it would be higher than it dims too much with very high FALL.

aron7awol

31st December 2020, 00:29

Reverse engineering FTW! :)

I'm confident I figured out the formula:

log10(DPL)*DTN/50*FALL and optionally +DPL

chros

31st December 2020, 00:39

:D That's what I meant!
But do you think that DPL is still in there???

aron7awol

31st December 2020, 01:44

:D That's what I meant!
But do you think that DPL is still in there???
Huh? I just posted the formula, did you miss it? log10(DPL) is in there.

So essentially at 1000 nits we are getting an extra 1.5x multiplier over 100 nits. This is fine, we can adjust DTN for this.

Multipliers at 1000 nits:
DTN 33 = 1.98xFALL, similar to 100 nits and DTN 50
DTN 37 = 2.22xFALL, similar to 100 nits and DTN 55
DTN 40 = 2.40xFALL, similar to 100 nits and DTN 60
DTN 43 = 2.58xFALL, similar to 100 nits and DTN 65
DTN 47 = 2.82xFALL, similar to 100 nits and DTN 70
DTN 50 = 3.00xFALL, similar to 100 nits and DTN 75

DPL is still very much involved but at least it's no longer being added to it on top, and so we can adjust for it.

quietvoid

31st December 2020, 19:32

Went back to tweaking my DTN again, I can't go any lower than 85 for the extreme FALL cases in the S&M demo. Otherwise there's a bunch of washed out highlight detail.
Looked quickly at Fury Road and the new saturation option seems to be really good at retaining luminance.

I'm gonna settle with these settings for now: https://ibb.co/HtSyV6W

aron7awol

31st December 2020, 23:37

Went back to tweaking my DTN again, I can't go any lower than 85 for the extreme FALL cases in the S&M demo. Otherwise there's a bunch of washed out highlight detail.
Which scenes specifically force you to go that high with DTN? I'd like to test with them.

Also, which scenes made you choose desat 3 2000? I don't think we ~1000 nit people need nearly as much desat as those ~100 nit guys. Don't you find that strong desat 3 turns bright highlights too white? I just tested the green spear and MMFR explosions with your settings and it seemed like too much desat to me. Obviously this is preference.

quietvoid

1st January 2021, 00:38

I look at the cloud scenes mostly, 1:18 clouds and 2:26 rocks lose a bit too much for my liking. There's still loss in the "wave hits rocks" one, but that's gonna have to be a compromise of lowering for more overall brightness.
85 is not very high compared to what we had at 150 before!

For saturation I only looked at some flames and lightning in Fury Road and the sunset scene in S&M around 2:09 (high saturation in the clouds).
There's a bit more detail visible after desaturating, I didn't really compare lower strength so maybe 2000 is too much but at least it's not flat red explosions.

I guess it goes against the coded values for Fury Road, but I do prefer lightning and flames looking slightly whiter than blue-ish, because it increases perceived brightness.
I'll look at the green spear now.

Edit: Ah well madshi made another build that I'll have to readjust for :)

quietvoid

1st January 2021, 02:52

Had a quick look with 117, not sure I'd go over 150 at all. Probably going desat 3, 125.

SamuriHL

1st January 2021, 04:04

We need to play with the ceiling multiplier and find out what works for us in 117. Then we can have that option permanently.

quietvoid

1st January 2021, 04:11

I tried it but I don't know in what type of scenes it's supposed to have an effect.
Couldn't see any difference so far.

SamuriHL

1st January 2021, 05:50

Very bright scenes.

aron7awol

1st January 2021, 06:37

What I did for my test was I set one profile to disable ceiling and used that one to hone DTN, and then I set another profile to a really high DTN like 250 but with the ceiling in place and use that one to hone the ceiling multiplier. Then as I found values I liked for each (150 and 4x), going through each scene and toggling between the two, I would find some scenes DTN resulted in a lower target (pretty high FALL but many pixels <100 nits), and in others the ceiling would result in a lower target (when almost all pixels are >100 nits), and in each case I preferred the lower target.

chros

1st January 2021, 13:13

@chros Since the CX seems to have received an update that fixes the Dolby Vision black levels, I digged through posts to find another test pattern: https://www.avsforum.com/threads/spears-munsil-uhd-hdr-benchmark-disc-discussion.3075780/page-43#post-59634582
Plays through the internal player.
On my set there's no discernible difference starting at 72, and then lower bars blend in with the non black background, at brightness 50.
When lowering to 49 I get absolute black but bars below 83 disappear. :)

Seems to be perfectly aligned with the 0 - 2.3% clipping, at least!
Now I don't know which compromise to pick.

:) Same here from the Oppo. Well, HDR10 is slightly better in PC mode with 76 visible, so dunno. Probably I'll leave it as it is.

So, I played with this a bit, and there's another option that works completely differently, but don't get too excited, I went back to the default settings (b50, c100) :)

Option 1: b49, c100
- in real scenes, the image remains vivid, no color hue (?) shift and that is good, but the whole image gets darker (not just the low end)!
-- it works like the "double/triple expanded" of "source level" toggle of madvr! (and we just lower the brightness with 1 step! :D)
- overall, for me, the image gets unnatural, I couldn't get used to it

Option 2: b50, c99
- yes, that's right, lowering contrast just by 1 results in perfect black using the 2 DV patterns we have: 76 is visible (vs 73 and 83)
-- this option works in a weird way, in every 3 steps (97/94/etc) it has a "glitch"
- I also tested the higher end (with Oppo forcing DV with Mehanik's HDR10 patterns) and there was nothing outstanding and that is good
- but when I thought we found the solution for this, I tested with real scenes (when watching content) :D :
-- the image has less pop/vivid, becomes a bit dull
-- there are color hue (?) shift, visible almost everywhere (skins, background, etc)
-- but the whole image does not get darker! this setting only affects the low end!

So, as far as I see, we are out of options: we have to "wait" for LG to get their **** together :D

@aron, have you received the DV fix firmware for your CX? If so, would you mind testing it with default Cinema (not home) preset? (not urgent)

chros

1st January 2021, 13:42

I just posted the formula, did you miss it? log10(DPL) is in there.
...
DPL is still very much involved but at least it's no longer being added to it on top, and so we can adjust for it.
I quickly compared to my results (https://forum.doom9.org/showthread.php?p=1931951#post1931951) with b115, and it looks that way indeed.
But did he removed the DPL from the formula for b117? :)

I tried it but I don't know in what type of scenes it's supposed to have an effect.
Couldn't see any difference so far.

What I did for my test was I set one profile to disable ceiling and used that one to hone DTN, and then I set another profile to a really high DTN like 250 but with the ceiling in place and use that one to hone the ceiling multiplier. Then as I found values I liked for each (150 and 4x), going through each scene and toggling between the two, I would find some scenes DTN resulted in a lower target (pretty high FALL but many pixels <100 nits), and in others the ceiling would result in a lower target (when almost all pixels are >100 nits), and in each case I preferred the lower target.
I agree with everything you said (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60378607), and that's what I tried to tell (https://forum.doom9.org/showthread.php?p=1930893#post1930893) the others as well:
"And that's also a good example that S&M is not the go-to clip to check, it's not representative at all (more like an extreme case), it's just another content to check"

We have to know what the used formula is (!), what the definition of highlights is (!), etc, we can't just faq around in the dark :D It won't work that we (!) try to come up with a valid solution without knowing anything about how it works ...
*Or* we have to come with our own formula ...
You already asked both from him, did you get an answer?! ...
And I even started to doubt he tests things ... especially with HDR output ...

Anyway, this new "don't add peak nits" option is already a big step for us, especially with 700 DPL (with 1000 DPL you are not affected close as much).

chros

1st January 2021, 14:05

But did he removed the DPL from the formula for b117? :)
He didn't, after quick testing.

aron7awol

1st January 2021, 16:48

The only change in the formula for b117 is whether DPL is added, the "optionally" section below. The formula I figured out is the formula in use:

Reverse engineering FTW! :)

I'm confident I figured out the formula:

log10(DPL)*DTN/50*FALL and optionally +DPL

So yes, DPL is still in there, but at least it's part of the multiplier and thus easy enough to adjust for, as described in my post about 10 posts up. It's an extra 1.5x multiplier at 1000 nits vs 100 nits (3/2, because 1000=10^3 and 100=10^2).

SamuriHL

2nd January 2021, 03:22

aron7awol

2nd January 2021, 18:29

- yes, that's right, lowering contrast just by 1 results in perfect black using the 2 DV patterns we have: 76 is visible (vs 73 and 83)
-- this option works in a weird way, in every 3 steps (97/94/etc) it has a "glitch"
- I also tested the higher end (with Oppo forcing DV with Mehanik's HDR10 patterns) and there was nothing outstanding and that is good
- but when I thought we found the solution for this, I tested with real scenes (when watching content) :D :
-- the image has less pop/vivid, becomes a bit dull
-- there are color hue (?) shift, visible almost everywhere (skins, background, etc)
-- but the whole image does not get darker! this setting only affects the low end!
So, not even related to DV, just regular HDR10, I noticed this behavior when changing contrast from 100 to 99 while I was testing near-black gradation on my CX. It has a substantial effect on the near-black gradation, and I did notice the effect sort of "cycles" every few clicks like you noticed. And I also noticed that dropping from 100 to 99 there was a major hue shift everywhere. Interestingly, this shift at 99 actually made the hue more closely match my Sony X900F that I was comparing side-by-side.

quietvoid

2nd January 2021, 21:15

Played with 118 and the new ceiling option.
I've switched to 150 + 4x as well, it retains more contrast and detail while not making anything unnecessarily dimmer.
Does better than my 85 DTN + disable checked, which would clip much more in flashes like in MMFR.

For saturation, I toned it down a bit from 50/125 to 75/125 (higher mixture seems to lower desaturation at the same strength).
I don't know why some of you have disabled the fire/explosions tweak, it makes explosions pink-ish and I don't really like that.

Set the new repair stuff to what was mentioned as good on AVS, my new settings: https://ibb.co/cwQzGKj

chros

2nd January 2021, 22:46

Just a note, you set no compr limit to 203, I think it doesn't do anything (since it is smaller than DPL). That's why set it to 0, not to confuse everyone.

About other settings: I use 700 DPL vs 1000. That's a big difference for me, here's why.
Most of the content today (95% ?) has <=1000 maxCLL metadata (that sometimes not perfect, but close enough), even DV releases! That means for you that madvr doesn't have to compress anything, DTN doesn't work! And that's with 95% of the content! :)

Now, with 700 DPL, even with the contents above, DTN will kick in sometimes. If I use 60 DTN, it won't have an effect until ~207 FALL, which sounds reasonable with 1000 maxCLL content.

And if we keep thinking this way, I doubt that there are many movies (e.g. 10 out of 1000 = 1%?) that has content shifted up. I even argue that they are bad master! If that's the case then probably they need separate profile, and all the "normal" content should use "conservative" settings.

Try to find "normal" movies/scenes with >1000 nits and reasonably high FALL (>110?) to verify your current settings, I think it will fail.
(With 1000 DPL and 150 DTN every frame that has ~>110 FALL will unnecessarily dim. This FALL value with 700 DPL and 150 DTN is 83! (vs 110))

quietvoid

2nd January 2021, 22:51

I'm aware the no compr limit does nothing.

About DPL, it's noticeably dimmer with 700, there's no more detail resolved at all, and 1000 doesn't hurt at all.
Like some mentioned before, the C8 (or whatever) can do over 700 nits for specular highlights, so using 700 is limiting.

I don't really remember the other arguments, but I think the curve used also follows PQ better, while 700 rolls off early.

chros

2nd January 2021, 23:16

I know, but I have a B8 :) that only reaches 700 and not 900ish with your C8 (Sony Oleds only reach 600 tops). So I just wanted to make a point that for me with B8 the settings are different. (I edited a bit the post above, I'd be curious about your thoughts/findings)

chros

2nd January 2021, 23:40

I don't really remember the other arguments, but I think the curve used also follows PQ better, while 700 rolls off early.
The curve is the same for sure between 840 and 1000 (maybe even from 801).
I don't want to say that you can't use 1000 DPL (but I don't think it's valid, due to ABL with real scenes). And Aron's CX is dimmer, maybe 800 peak?

Magik Mark

2nd January 2021, 23:44

QuitVoid, Your setting is just perfect for my C9. Thanks for sharing. Other settings that I have tried is also good but a little bit dim especially if you don't have a dedicated HT Room

quietvoid

3rd January 2021, 00:22

The curve is the same for sure between 840 and 1000 (maybe even from 801).
I don't want to say that you can't use 1000 DPL (but I don't think it's valid, due to ABL with real scenes). And Aron's CX is dimmer, maybe 800 peak?

Well maybe it's not necessary to go lower, since the TV seems capable of mapping 1000 peak pretty well.
I'd have to compare 800-900-1000 to know more, but the weird curve up to 800 definitely started clipping above 760.

I guess that's next, maybe others have done this already and can share.

I'm not too sure about what you mean with FALL, are you saying you have dimming when over a certain FALL even when the peak brightness is under DPL?
I don't know many scenes to test this with otherwise.

chros

3rd January 2021, 00:36

No, I say that it will dim for you with certain FALL (>110) and measured peak above (!) DPL.

aron7awol

3rd January 2021, 00:38

About other settings: I use 700 DPL vs 1000. That's a big difference for me, here's why.
Most of the content today (95% ?) has <=1000 maxCLL metadata (that sometimes not perfect, but close enough), even DV releases! That means for you that madvr doesn't have to compress anything, DTN doesn't work! And that's with 95% of the content! :)

Now, with 700 DPL, even with the contents above, DTN will kick in sometimes. If I use 60 DTN, it won't have an effect until ~207 FALL, which sounds reasonable with 1000 maxCLL content.
Well, setting it to 700 if the display can't go above that makes total sense to me. :) Otherwise, I don't want DTN to do anything just for the sake of doing something. In other words, on a display capable of 1000 nits, I don't want DTN to do anything on this ~95% of films.

Try to find "normal" movies/scenes with >1000 nits and reasonably high FALL (>110?) to verify your current settings, I think it will fail.
(With 1000 DPL and 150 DTN every frame that has ~>110 FALL will unnecessarily dim. This FALL value with 700 DPL and 150 DTN is 83! (vs 110))
I agree, 150 DTN is really high, that results in a 9x FALL multiplier with 1000 DTN!

quietvoid

3rd January 2021, 00:47

aron7awol

3rd January 2021, 00:50

About DPL, it's noticeably dimmer with 700, there's no more detail resolved at all, and 1000 doesn't hurt at all.
Like some mentioned before, the C8 (or whatever) can do over 700 nits for specular highlights, so using 700 is limiting.
I found the same in my quick testing of 700 vs 1000 DPL. 700 is noticeably dimmer.

Now, I'm not sure what my display is actually capable of, but I definitely see small highlights getting brighter at 1000 DPL vs 700 or 800 or 900.

chros

3rd January 2021, 00:56

In other words, on a display capable of 1000 nits, I don't want DTN to do anything on this ~95% of films.
Yes :) but your CX can't do 1000 for sure either! Do you have a colorimeter?

Qietvoid, I'll try to find such an example for you to be able to see what I mean.

aron7awol

3rd January 2021, 00:57

Okay I get it. I did briefly test using 125 DTN but it would still clip some stuff.

Maybe we have different goals, I'd rather resolve more detail/contrast than preserve absolute brightness.
With the new avgHighlights ceiling at 4x, there's less dimming than with 150 DTN + ceiling disabled.
So this alone already reduces dimming without affecting detail.

It was equivalent to 85 DTN + ceiling disabled for some of the scenes, while being better for contrast in others.
There's no unnecessary dimming at all from the couple samples I tested.
Is that with S&M?

Definitely true on different goals, and I wouldn't want to try to tell you how you should run your system. :) But I agree with @chros that choosing our settings based on content that we will never encounter during normal usage (10,000 nit S&M, for example) will likely result in sub-optimal settings for the majority of content out there. I'm not saying that's what you did, I'm interested in finding out what content you did test with.

I think you tried the 150 and 4x combo based on my post in the AVS thread where I mentioned that's what I settled on for the 10,000 nit S&M demo. But since then, I tested The Meg, and liked 75-85 and 3x a lot better on that. So there is certainly no one-size-fits-all combo of settings, at least not with how the algos are currently constructed.

aron7awol

3rd January 2021, 01:06

Yes :) but your CX can't do 1000 for sure either! Do you have a colorimeter?
I know, I was careful to say "a display capable of 1000 nits", not "my display capable of 1000 nits". :)

I do have a colorimeter, but I haven't tried to measure peak luminance yet. What window sizes have you measured?

The way I look at it, in a perfect world, we would measure what our display is capable of in a tiny specular highlight and set DPL to that. Then in the same perfect world, if a larger highlight occurs that it can not produce at that same peak luminance, ABL kicks in and dims the whole picture slightly, just as DTN does.

quietvoid

3rd January 2021, 01:06

Is that with S&M?

Definitely true on different goals, and I wouldn't want to try to tell you how you should run your system. :) But I agree with @chros that choosing our settings based on content that we will never encounter during normal usage (10,000 nit S&M, for example) will likely result in sub-optimal settings for the majority of content out there. I'm not saying that's what you did, I'm interested in finding out what content you did test with.

I think you tried the 150 and 4x combo based on my post in the AVS thread where I mentioned that's what I settled on for the 10,000 nit S&M demo. But since then, I tested The Meg, and liked 75-85 and 3x a lot better on that. So there is certainly no one-size-fits-all combo of settings, at least not with how the algos are currently constructed.Yes S&M, I did look at some shots from MMFR but that too is a bit extreme :D
Though there's less "visible" detail in MMFR (mostly just really bright white blobs), the contrast is increased.

I used to do the old 100 DTN + ceiling disabled for The Meg, might have to look at that again.

chros

3rd January 2021, 13:40

Definitely true on different goals
I disagree :) all we have the same goal, but we don't see it yet :)

Is that with S&M?
Yes S&M
My only point is this:
I agree, 150 DTN is really high, that results in a 9x FALL multiplier with 1000 DTN!
But I agree with @chros that choosing our settings based on content that we will never encounter during normal usage (10,000 nit S&M, for example) will likely result in sub-optimal settings for the majority of content out there.

I used to do the old 100 DTN + ceiling disabled for The Meg, might have to look at that again.
Plus, I think even The Meg and Fury Road are badly mastered, so these are not representative either.

Qietvoid, I'll try to find such an example for you to be able to see what I mean.
Check out The Grinch (2018): although it has 1000 maxCLL, the content goes above it many times, and there are plenty of scenes >100 FALL, even above 300 FALL.
This is also a DV FEL title, so you can compare it if you like. Let me know if you don't like this title for whatever reason.
I'd be curious what you think after testing your settings with this. :)
Because I think you'd rather encounter such scenes way more often than The Meg, MMFR, SM 10000 demo :)

So there is certainly no one-size-fits-all combo of settings, at least not with how the algos are currently constructed.
Exactly! And who knows, maybe future curves will behave differently.

I do have a colorimeter, but I haven't tried to measure peak luminance yet. What window sizes have you measured?
Can you measure it with 1% window and 20 point greyscale? See the HCFR settings (https://www.avsforum.com/threads/2018-lg-oled-calibration-and-user-settings-no-price-talk.2962814/post-58739264) at the beginning of my B8 post and attached HCFR screenshot below the post (enable values on the graph, right click on the graph and you can zoom in/out, etc). Can you upload the graphs after that? Thanks
Each run should be under 2 minutes, let the display warm up for 10 mins at first, then let 2-5 mins cool down period between runs.
Can you do at least 3 runs with the following metadata?
- 1. 1000, 1000, 333
- 2. 840, 840, 280
- 3. your measured peak from run 1. and 2. (should be the same), maxFALL should be maxCLL/3

The first 2 should result in the very same curve! (at least this is with my B8)

Here's rtings's result (https://www.rtings.com/monitor/reviews/lg/48-cx-oled), they state that HDR 2% window peak is 725! :) So your result should be around this (+-30-50 nits but not more.)
And their HDR real scene peak result is: 571 nits! They used this clip (https://i.rtings.com/images/test-materials/2017/HDR.mp4): it has 2 big white moving patches, no metadata, 10.000 nits CLL (!) and very high 340 FALL! So, not sure how valid this is :)

Edit: and they have another CX page as well, now with Cinema preset (https://www.rtings.com/tv/reviews/lg/cx-oled#comparison_5668), result in this case:
- 2% window: 799
- HDR Real Scene Peak: 685
What interesting with this test, that they got higher peak with 10% window than 2%! :) (799 vs 813)

I'm asking for this test, because no one else did this apart from me. @QBhd tried it with his C8 but he stated that he didn't get reliable result, that I don't understand, because I got the (almost) the same every time I did it.
Cheers

The way I look at it, in a perfect world, we would measure what our display is capable of in a tiny specular highlight and set DPL to that. Then in the same perfect world, if a larger highlight occurs that it can not produce at that same peak luminance, ABL kicks in and dims the whole picture slightly, just as DTN does.
I agree.

@SamuriHL, do you know whether the ColorSpace guys did any measurements like this? If not, can you ask them what they think about this? 1% window vs who-knows-what vs ABL? :) (I haven't read that topic for half a year now) Thanks

I found the same in my quick testing of 700 vs 1000 DPL. 700 is noticeably dimmer.

Now, I'm not sure what my display is actually capable of, but I definitely see small highlights getting brighter at 1000 DPL vs 700 or 800 or 900.
Now that's interesting, since both of you mentioned this!
Usual question: with which source/scenes? :) S&M 10000? :) And even at DPL 900 vs 1000?
I noticed somewhat similar with SM crocodile scene (there's not a big difference here), but I'm 99% sure that's the ABL with 700 DPL due to really high FALL.

Can you guys see a similar effect e.g. with John Wick2 sample (lights on the firetruck at the beginning)?

What do you think, let's focus on trying to find out what valid DPL we can use (until others fool around with desaturation :) )?

aron7awol

3rd January 2021, 16:02

I disagree :) all we have the same goal, but we don't see it yet :)
There is definitely plenty of personal preference on prioritization of highlight detail, overall brightness, an acceptable amount of clipping, etc. I don't see how everyone could possibly ever agree on all of that.

Edit: and they have another CX page as well, now with Cinema preset (https://www.rtings.com/tv/reviews/lg/cx-oled#comparison_5668), result in this case:
- 2% window: 799
- HDR Real Scene Peak: 685
What interesting with this test, that they got higher peak with 10% window than 2%! :) (799 vs 813)
I did see that they measured ~800 nits on 2% and 10%, but I wondered if that would increase on even smaller windows.

I'm asking for this test, because no one else did this apart from me. @QBhd tried it with his C8 but he stated that he didn't get reliable result, that I don't understand, because I got the (almost) the same every time I did it.
I'd love to do this test and gather the data, but there are so many things I should be doing instead of even posting on these forums about TMing, so I don't see myself taking the time to do this anytime soon unfortunately!

Especially since ultimately, these measurements are all well and good and valuable to have, but I'm gong to let the real-world results guide my decision-making on this. So if 800 or 900 or 1000 or something in between gives the overall best perceived result, I'm going to go with that. I think that's the best approach given the limited control we have over the display.

Now that's interesting, since both of you mentioned this!
Usual question: with which source/scenes? :) S&M 10000? :) And even at DPL 900 vs 1000?
I did some quick testing of this stuff last night just jumping around to a bunch of different movies and scenes, some overall bright and some overall dark with small bright highlights. One example is in Tomb Raider when they are underground with bright flashlights. On that one, there was definitely a noticeable and beneficial change going from 800 to 850 to 900, but not so much from 900 to 1000. I didn't go into more detail and check some things in between (875? 950?). I also remember on the green spear the difference I noticed between 900 and 1000 seemed to be maybe some clipping of the details (the halo around the spear became more pronounced and flatter on 1000). So based on that quick testing last night, if I had to pick something right now, I'd prob pick 900. I think that's probably close to where I'll end up, but this certainly needs more testing.

I noticed somewhat similar with SM crocodile scene (there's not a big difference here), but I'm 99% sure that's the ABL with 700 DPL due to really high FALL.
Are you saying you think the difference there was ABL kicking in differently at 700 vs higher?

chros

3rd January 2021, 16:34

I did see that they measured ~800 nits on 2% and 10%, but I wondered if that would increase on even smaller windows.
Probably, just a bit. But their 10% window result is really strange...

I'd love to do this test and gather the data, but there are so many things I should be doing instead of even posting on these forums about TMing, so I don't see myself taking the time to do this anytime soon unfortunately!
Welcome to the club! :) No worries, when you're in the mood... (and have 30 mins)

I did some quick testing of this stuff last night just jumping around to a bunch of different movies and scenes, some overall bright and some overall dark with small bright highlights. One example is in Tomb Raider when they are underground with bright flashlights. On that one, there was definitely a noticeable and beneficial change going from 800 to 850 to 900, but not so much from 900 to 1000. I didn't go into more detail and check some things in between (875? 950?). I also remember on the green spear the difference I noticed between 900 and 1000 seemed to be maybe some clipping of the details (the halo around the spear became more pronounced and flatter on 1000). So based on that quick testing last night, if I had to pick something right now, I'd prob pick 900. I think that's probably close to where I'll end up, but this certainly needs more testing.
Interesting, thanks. Which Tomb Raider is that (there are at least 3), timestamp?
Do you have a link for the green spear or can you upload it?

Are you saying you think the difference there was ABL kicking in differently at 700 vs higher?
That's correct. 700 curve is much brighter and that can easily trigger ABL with high FALL.

aron7awol

3rd January 2021, 18:44

Interesting, thanks. Which Tomb Raider is that (there are at least 3), timestamp?
The newest Tomb Raider with Alicia Vikander. She drops into the tomb at 1:18:30 and then there are a whole bunch of dark scenes with bright flashlights and torches from then on.

Do you have a link for the green spear or can you upload it?
I don't, but I just asked in the AVS thread if anyone has it.

That's correct. 700 curve is much brighter and that can easily trigger ABL with high FALL.
Hmm, maybe the best way to test this then is to use DTN settings that end up at the same target?

For example:
DPL 953, DTN 79
DPL 875, DTN 80
DPL 805, DTN 81
DPL 742, DTN 82

chros

3rd January 2021, 19:15

aron7awol

3rd January 2021, 19:33

About last point: but do we know that they end up with the same lightness? (who knows what madvr curve does at different DPLs) So not sure it's a valid test either.
Cheers, I'll take a look at TR with 700-800-900 DPL. Btw this is also a DV title, MEL, so we can try to compare it with madvr (not that DV should be the etalon :)).
It's still not perfect, but I think it's much better than comparing them the other way.

If we just compare DPL of 742 vs DPL of 875 on the same scene without DTN, we are comparing two different TMing curves.

If we instead compare them with the corresponding DTN settings I described, and compare a frame where DTN kicks in and they end up with the same target, it is the same TMing curve, the only difference is then what the result is scaled down to. This also much more closely matches what will actually happen when we run madVR with these different settings.

Testing it this way on BvS I could see some detail getting clipped a bit at 953 vs 875.

aron7awol

3rd January 2021, 20:34

Here are some pairings more appropriate for you:

DPL 700, DTN 60
DPL 782, DTN 59
DPL 877, DTN 58

SamuriHL

3rd January 2021, 20:34

I've run out of vacation time so I don't have time to do any more testing right now. I did measure my 1% window with colourspace a few weeks back and posted my results. I recall the 1% window being in the 900 nit range but completely unstable as the panel heats up and pushes to the point where ABL kicks in and it starts to drop again. Realistically if we can guarantee that at 900 (or less for panels that can't reach that high) the tone curve is the same as 1000 on these OLED panels, I'd be in favor of dropping DPL. I WON'T go back to 700 after what we saw in the 501->xxx range. I never checked the high end of that crazy nonsense since I just set it to 1000 and called it good. But the more I think about it, setting it to whatever you measure for your panel makes a lot of sense so long as the tone curve is correct on the display. This, btw, is why I kept asking for a way to set the metadata to 1000 while setting a real DPL. You guys told me that was insane and not needed and why the hell is he asking for that. Now you know. :) I want to be able to accurately set the DPL and force the display to use the right tone curve to display it.

aron7awol

3rd January 2021, 21:02

Given the limited control we have over the displays' TMing, I think these are the things we have to balance:

1. We don't want to set DPL too low, because madVR will never (I don't think) output a pixel > DPL.
2. We don't want to set DPL too high, because we will start to clip highlights, as madVR starts asking for pixels to be brighter than the display will actually produce. Or depending on preference, we want to clip, but only to some desired extent.
3. The display's curve changes based on the metadata, which (currently) changes with DPL.

So I totally agree on wanting to be able to override metadata, and I asked madshi to add this functionality to madVR. Because then we could set DPL to what we want based on 1 and 2, without having to consider 3.

Then @chros, for example, might decide to set DPL to ~680 and set the metadata to 840 or 960, if he wants to avoid clipping as much as possible but squeeze as much as he can out of the possible curves.

SamuriHL

3rd January 2021, 21:57

Exactly. *I* can already do this. :p But it's a pain in the ass.

chros

4th January 2021, 12:28

Cheers, I'll take a look at TR with 700-800-900 DPL. Btw this is also a DV title, MEL, so we can try to compare it with madvr (not that DV should be the etalon :)).
I tried to spot any major visible difference around 1:18:00 (scenes can be as high as 3000 nits), toggling between 700/840/1000 DPL, although there was some small diff in the light of the torches but honestly I would fail in a blind test miserably :) (Obviously discounting the brighter image at 700 DPL). And the diff also can come from the fact that madvr has to compress more with 700/840 than e,g, 1000.

I fired up the DV version, but not having side-by-side screens, again, I wouldn't know which is which (discounting the issues with HDR10 on LG Oled).
So, not sure what conclusion I can draw from this. :)

If you come across with scenes where the difference if obvious then let us know.

If we instead compare them with the corresponding DTN settings I described, and compare a frame where DTN kicks in and they end up with the same target, it is the same TMing curve, the only difference is then what the result is scaled down to. This also much more closely matches what will actually happen when we run madVR with these different settings.

Here are some pairings more appropriate for you:

DPL 700, DTN 60
DPL 782, DTN 59
DPL 877, DTN 58

Cheers, I'll take a look at the crocodile scene of SM 10000.

I did measure my 1% window with colourspace a few weeks back and posted my results. I recall the 1% window being in the 900 nit range
I don't remember, maybe I have to drink more :D Was it in this tread? If so I'll look it up.

Realistically if we can guarantee that at 900 (or less for panels that can't reach that high) the tone curve is the same as 1000 on these OLED panels, I'd be in favor of dropping DPL. I WON'T go back to 700 after what we saw in the 501->xxx range.
This is the last time when I tell you guys, that the curve is the same between 840-1000 for sure!!! :) (At least on 2018 models.)

But you don't have to measure this (if you don't want to):
- select a normal bright scene with almost 800 nits
- toggle between 840/900/1000 DPL and you won't see any difference!

About the 501-800 range: this can be good for sunny daytime viewing, instead of turn on LG's DTM, we can set e.g. 600 and mnadvr will deal with compression and the APL will be brighter due to the curve is used vs e.g. 900.

So, due to our discussion about this, yesterday evening I started to set 840 DPL for night viewing (to have almost proper tone curve), and will use 700 for daytime viewing. I'll see whether I can spot anything unusual (although I doubt it :)).

I never checked the high end of that crazy nonsense since I just set it to 1000 and called it good. But the more I think about it, setting it to whatever you measure for your panel makes a lot of sense so long as the tone curve is correct on the display.
I'm still not sure about this, maybe you can go higher as you guys suggested, dunno ...
Btw, yesterday I quickly tried to measure the spotlights in TR scenes: no chance :) The highest I could measure was 430 nits, and that's nowhere near to the 2000/3000 nits that madvr reported. So who knows where are these specular highlights are and how bright they can actually be on the display.

This, btw, is why I kept asking for a way to set the metadata to 1000 while setting a real DPL. You guys told me that was insane and not needed and why the hell is he asking for that. Now you know. :) I want to be able to accurately set the DPL and force the display to use the right tone curve to display it.
In your case it's easy, if you measured 900, you can set it, and all is good for you. But that's not true for me with the B8 (although I'm not bothered about this).

But we all were agreed that it's a useful feature, all of us asked madshi to implement it (I even told him to leave the useless split screen), what happened? :) He didn't do anything about this but implemented the split screen for SDR (!) for pj users, fooling around 50 with 100 DPL :D

Given the limited control we have over the displays' TMing, I think these are the things we have to balance:

1. We don't want to set DPL too low, because madVR will never (I don't think) output a pixel > DPL.
2. We don't want to set DPL too high, because we will start to clip highlights, as madVR starts asking for pixels to be brighter than the display will actually produce. Or depending on preference, we want to clip, but only to some desired extent.
3. The display's curve changes based on the metadata, which (currently) changes with DPL.

So I totally agree on wanting to be able to override metadata, and I asked madshi to add this functionality to madVR. Because then we could set DPL to what we want based on 1 and 2, without having to consider 3.

Then @chros, for example, might decide to set DPL to ~680 and set the metadata to 840 or 960, if he wants to avoid clipping as much as possible but squeeze as much as he can out of the possible curves.
Agreed, we have to ask him countless times and maybe one day he will implement it :)

I've run out of vacation time
You don't have to tell me, I'm already in the "mine", "pushing the carts" :D

SamuriHL

4th January 2021, 15:26

I don't remember, maybe I have to drink more :D Was it in this tread? If so I'll look it up.

It was indeed. I don't remember the exact numbers I measured off the top of my head but it was pushing close to 900 nits. But man it fluctuates wildly up there as you would expect on these panels.

This is the last time when I tell you guys, that the curve is the same between 840-1000 for sure!!! :) (At least on 2018 models.)

But you don't have to measure this (if you don't want to):
- select a normal bright scene with almost 800 nits
- toggle between 840/900/1000 DPL and you won't see any difference!

About the 501-800 range: this can be good for sunny daytime viewing, instead of turn on LG's DTM, we can set e.g. 600 and mnadvr will deal with compression and the APL will be brighter due to the curve is used vs e.g. 900.

So, due to our discussion about this, yesterday evening I started to set 840 DPL for night viewing (to have almost proper tone curve), and will use 700 for daytime viewing. I'll see whether I can spot anything unusual (although I doubt it :)).

When I get more time I was going to try 900 and see what that looks like. I suspect it'll be just fine and for my C8 it should be a decent value to set. Daytime viewing of HDR is illegal, btw. :p :D LMAO!

I'm still not sure about this, maybe you can go higher as you guys suggested, dunno ...
Btw, yesterday I quickly tried to measure the spotlights in TR scenes: no chance :) The highest I could measure was 430 nits, and that's nowhere near to the 2000/3000 nits that madvr reported. So who knows where are these specular highlights are and how bright they can actually be on the display.

I thought madvr used the 99% measurement so that if you had, e.g. one pixel (you get the idea) that was super bright it'd not impact the entire frame? That would suggest the frame you're measuring has at least a bit more than 1% of high nit pixels. Somewhere. LOL

In your case it's easy, if you measured 900, you can set it, and all is good for you. But that's not true for me with the B8 (although I'm not bothered about this).

But we all were agreed that it's a useful feature, all of us asked madshi to implement it (I even told him to leave the useless split screen), what happened? :) He didn't do anything about this but implemented the split screen for SDR (!) for pj users, fooling around 50 with 100 DPL :D

ROFLMAO! Yea I don't know that I could live with "HDR" at 50 nits. How the hell that even works is beyond me. And expecting it to look identical to a 10000 nit screen seems..."unrealistic". lol

Agreed, we have to ask him countless times and maybe one day he will implement it :)

You don't have to tell me, I'm already in the "mine", "pushing the carts" :D

Yea I'm about to join my first meeting since the 23rd. UGH LOL

aron7awol

4th January 2021, 15:43

I tried to spot any major visible difference around 1:18:00 (scenes can be as high as 3000 nits), toggling between 700/840/1000 DPL, although there was some small diff in the light of the torches but honestly I would fail in a blind test miserably :) (Obviously discounting the brighter image at 700 DPL). And the diff also can come from the fact that madvr has to compress more with 700/840 than e,g, 1000.

I fired up the DV version, but not having side-by-side screens, again, I wouldn't know which is which (discounting the issues with HDR10 on LG Oled).
So, not sure what conclusion I can draw from this. :)

If you come across with scenes where the difference if obvious then let us know.

Cheers, I'll take a look at the crocodile scene of SM 10000.

At those DPLs, and with your measured display peak, the only difference you might be seeing there is the slightest bit of clipping.

The scene that I found to show this clipping a lot better was in BvS when Doomsday "goes nuclear" or whatever you would call it. I don't know the timestamp off the top of my head. In the explosion I could see differences between the different DPL/DTN pairings.

Btw, yesterday I quickly tried to measure the spotlights in TR scenes: no chance :) The highest I could measure was 430 nits, and that's nowhere near to the 2000/3000 nits that madvr reported. So who knows where are these specular highlights are and how bright they can actually be on the display.

Could this be because the meter is taking a reading that includes a lot of the surrounding dark pixels? So, say, the ~3000 nit flashlight only takes up about 15% of what the meter is capturing, so it reports ~450 nits? I admit I don't know how this works normally, just speculating. If so, relative measurements would still mean something, but then clipping would affect that so still not that valuable info necessarily.

I think the BvS scene is the best to show the clipping, and then let your eyes be the judge.

If you want to throw in an additional level that we know is below your real peak:

DPL 629, DTN 61
DPL 700, DTN 60
DPL 782, DTN 59
DPL 877, DTN 58

chros

4th January 2021, 19:10

I don't have the BvS movie (it's crap like many others :)), I need a sample for that.

aron7awol

4th January 2021, 19:49

Okay, I'm sure we can find another clip that shows a similar difference.

BTW, I sent you a PM on AVS regarding trying to do that peak nits testing you wanted me to do.

chros

4th January 2021, 21:10

Ok, I'll check it out. But can't you cut out a sample with mkvtoolnix and upload it?

aron7awol

4th January 2021, 22:19

Ok, I'll check it out. But can't you cut out a sample with mkvtoolnix and upload it?
Sure, I can do that.

quietvoid

5th January 2021, 00:09

I've spent a while testing again. 1000 DPL is definitely not optimal, at least for my C8.
While the image is overall brighter for high FALL scenes, highlights are either clipped or lose contrast (look flatter).

There's not much noticeable brightness loss at 900, but for me it's worth the extra detail/contrast in highlights.
To recover the brightness, I lowered DTN to 100 because it's more natural for those sky shots in The Meg, doesn't lose any detail with 900 DPL now.

With the BvS spear, 1000 also just looks like a green blob instead of being a more defined spear head.
840 makes it even more detailed but that also makes everything else dimmer, so I'm sticking with 900.
No benefit to using 800.

With 900 DPL, saturation seems lower so I increased mixture, explosions in MMFR look better defined and don't lose much compared to no desat.
The sep lum method can be used to increase saturation, and fix the grey/blue in flames, but it decreases perceptible contrast in the explosions too.
So max for better source reproduction, sep to make the MMFR car flames white.

The repair stuff doesn't seem to affect max at all, and I'd probably use 100/0.010 for sep.

Anyways, that's enough testing for me for a while :D
New settings: https://i.ibb.co/WPMPg37/madvr-118-new.png

SamuriHL

5th January 2021, 01:12

900 is where I was going myself so good to hear it works for you. DTN I'm still screwing around with. Desat is also something I'm still messing with although I've settled on max a while ago. Maybe I'll get some time to do more testing one night this week but it won't be tonight.

aron7awol

5th January 2021, 01:57

@quietvoid Did you try it without "apply desat 2-4"? I think desat 1 alone gives plenty of desat at our nit levels.

quietvoid

5th January 2021, 03:24

@quietvoid Did you try it without "apply desat 2-4"? I think desat 1 alone gives plenty of desat at our nit levels.

Yes, I don't think it's enough. Explosion looks flat, spear is just green.

aron7awol

5th January 2021, 05:21

Here's the sample:
https://mega.nz/file/0M0XXQiZ#zL2-ZzlqHQJGA5j98pAMZjai4iy4qr-3hlWLPcBV0Mk

After testing with this clip yesterday I was going back and forth between these two of my combos:
DPL 875, DTN 80
DPL 805, DTN 81

There's just the slightest hint of clipping detectable on the 875/80 combo when paused and looking really closely at certain things, but I couldn't really decide 100% which I preferred. They are both great.

chros

5th January 2021, 18:03

It was indeed. I don't remember the exact numbers I measured off the top of my head but it was pushing close to 900 nits. But man it fluctuates wildly up there as you would expect on these panels.
Cheers, then I'll try to find it. Yes, it goes up rapidly if you leave the white patch there :)

Daytime viewing of HDR is illegal, btw. :p :D LMAO!
:D I won't watch e.g. The Expanse during bright daylight, for sure :D

I thought madvr used the 99% measurement so that if you had, e.g. one pixel (you get the idea) that was super bright it'd not impact the entire frame? That would suggest the frame you're measuring has at least a bit more than 1% of high nit pixels. Somewhere. LOL
Can be, I don't remember. But that's what the dynamic clipping option is for, so maybe not.

ROFLMAO! Yea I don't know that I could live with "HDR" at 50 nits. How the hell that even works is beyond me. And expecting it to look identical to a 10000 nit screen seems..."unrealistic". lol
:D Yeah, this "new" HDR world completely ruined the pj world, I don't know who wants to have it anymore if doesn't want to sit down the whole street in front of it! :D

The scene that I found to show this clipping a lot better was in BvS when Doomsday "goes nuclear" or whatever you would call it. ... In the explosion I could see differences between the different DPL/DTN pairings.
Here's the sample:
https://mega.nz/file/0M0XXQiZ#zL2-ZzlqHQJGA5j98pAMZjai4iy4qr-3hlWLPcBV0Mk

After testing with this clip yesterday I was going back and forth between these two of my combos:
DPL 875, DTN 80
DPL 805, DTN 81

There's just the slightest hint of clipping detectable on the 875/80 combo when paused and looking really closely at certain things, but I couldn't really decide 100% which I preferred. They are both great.
Cheers, I'll have a look.

Could this be because the meter is taking a reading that includes a lot of the surrounding dark pixels? So, say, the ~3000 nit flashlight only takes up about 15% of what the meter is capturing, so it reports ~450 nits?
It can be as well, but those TR flashlights are "huge" even on a 65" screen, the middle of them definitely has different color (can be seen when switching profiles), so who knows :)

If you want to throw in an additional level that we know is below your real peak:

DPL 629, DTN 61
DPL 700, DTN 60
DPL 782, DTN 59
DPL 877, DTN 58
Cheers, I'll give it a try. But for me problem is that 700 is brighter (let alone 629) than higher DPLs, so it makes it really hard to compare them "objectively", I'll try anyway.

1000 DPL is definitely not optimal, at least for my C8.
While the image is overall brighter for high FALL scenes, highlights are either clipped or lose contrast (look flatter).

There's not much noticeable brightness loss at 900, but for me it's worth the extra detail/contrast in highlights.
To recover the brightness, I lowered DTN to 100 because it's more natural for those sky shots in The Meg, doesn't lose any detail with 900 DPL now.
Hm, that's not good news for me :) Because if you have 900 real peak, then it means 800 for me ..
If you remember scenes, timestamps ... :)

With the BvS spear, 1000 also just looks like a green blob instead of being a more defined spear head.
840 makes it even more detailed but that also makes everything else dimmer, so I'm sticking with 900.
No benefit to using 800.
Yep, no reason to go below the real peak.
Do you have a sample for the BvS spear by any chance?

With 900 DPL, saturation seems lower so I increased mixture, explosions in MMFR look better defined and don't lose much compared to no desat.
The sep lum method can be used to increase saturation, and fix the grey/blue in flames, but it decreases perceptible contrast in the explosions too.
So max for better source reproduction, sep to make the MMFR car flames white.

The repair stuff doesn't seem to affect max at all, and I'd probably use 100/0.010 for sep.

Interesting.

Anyways, that's enough testing for me for a while :D
:D And now we can wait another year to have a new meaningful option in madvr :D

BTW, I sent you a PM on AVS regarding trying to do that peak nits testing you wanted me to do.
And I just replied once more.

Klaus1189

5th January 2021, 18:14

LG CX Firmware 03.21.09 Improves Dolby Vision but Darkens HDR Game Mode - Fix Coming! (https://www.youtube.com/watch?v=Q5h1cvPApTQ)

SamuriHL

5th January 2021, 18:26

I love how our 2018 OLED screens were called junk by madshi. LMAO :D That makes me feel OH so great. It's all well and good at the "new" OLEDs don't hue shift, BUT.....they have shit for luminance compared to our C8's! I thought ~800 nits peak (10% window) was "low" but seeing how these new OLEDs can't even do 700...makes me never want to upgrade even with my "junk hue shifting panel". LOL

aron7awol

5th January 2021, 18:36

Cheers, I'll give it a try. But for me problem is that 700 is brighter (let alone 629) than higher DPLs, so it makes it really hard to compare them "objectively", I'll try anyway.
Well that's what these combos should adjust for almost entirely, so I'm thinking you'll find that it is (almost completely) no longer the case if you use these combos on a scene where DTN kicks in and the target is equal in all cases.

The only (just barely) perceivable difference I saw was just that tiniest bit of clipping.

quietvoid

5th January 2021, 19:40

Hm, that's not good news for me :) Because if you have 900 real peak, then it means 800 for me ..
If you remember scenes, timestamps ... :)

In the S&M demo, the horses in the snow scene looks flatter with 1000 DPL compared to 900, that's probably the most noticeable one.
You can also compare the crocodile one at the end and look at the rock specular highlights, though you might lose contrast going from 700 to 800.
That one is also flatter with 1000 for me, loses contrast on the rock.

This is the only sample I have for the spear: https://mega.nz/file/8J0kyYoQ#UpKJMyeENawlIJp3pwCYrOGBfYxNkVplVTSry35uYyk
It's not the untouched source.

aron7awol

5th January 2021, 19:54

In the S&M demo, the horses in the snow scene looks flatter with 1000 DPL compared to 900, that's probably the most noticeable one.
You can also compare the crocodile one at the end and look at the rock specular highlights, though you might lose contrast going from 700 to 800.
That one is also flatter with 1000 for me, loses contrast on the rock.

Sorry to keep harping on this, but did you do this comparison with DTN on? If so, did you adjust DTN for the different DPLs to end up at the same target level? Was this on the 10000 nit S&M, or 4000/1000? I just think it's really important that we aren't comparing differences in madVR TMing, but only differences in how it is handled by the display.

quietvoid

5th January 2021, 20:04

Sorry to keep harping on this, but did you do this comparison with DTN on? If so, did you adjust DTN for the different DPLs to end up at the same target level? Was this on the 10000 nit S&M, or 4000/1000? I just think it's really important that we aren't comparing differences in madVR TMing, but only differences in how it is handled by the display.

I did those with DTN on and the same DTN value, though it would still be related to the display TM I assume.
I don't expect the output to lose contrast at higher DPL unless the display itself maps the very high end of the curve differently.
If this was a difference from madVR, I'd think the specular highlights would remain the same, except at a different brightness.

There's not much difference in the scenes I mentioned when adjusting DTN, although 900/100 and 1000/125 do look the same in the horses scene, 1000 still looks worse in others (The Meg for example).
Obviously my method was flawed, but I don't feel like retesting just yet :)

Oh, and S&M is the 10 000 nits version.

aron7awol

5th January 2021, 20:42

The problem with not adjusting DTN so that you end up with the same target is that the curve will absolutely be different, and so you'll be seeing that difference, which will be a much larger difference than the tiny clipping difference. And so then any conclusion you try to draw is invalid, as you are essentially comparing the exact same thing you would compare if you left DPL the same in both cases and only changed DTN, "which DTN setting do I like better?". Whereas the question we're trying to answer is which DPL we like better.

Please try my DTN adjustment approach, guys, it really does work!

1000/125 is equivalent to 897/127 and 807/129
1000/98 is equivalent to 933/99 and 871/100 and 815/101

(this is with "don't add peak nits" checked)

SamuriHL

5th January 2021, 20:44

What is the DTN for 900?

aron7awol

5th January 2021, 20:57

What is the DTN for 900?
Since we can only do integer DTNs, I've been choosing DPLs that end up with the exact same target. So in that first set of values, I chose 897 instead of 900.

I'm happy to throw some combos out for you if you give me your "default" DPL and DTN combo you want to compare to.

Don't worry about the fact that the DPLs don't necessarily end up exactly where you want them. After you do the apples-to-apples comparison and see the relative differences and hone in on which you like best, we can then do further tweaking to get to the exact DPL you really want.

As an example, the perceived difference between my personal 875/80 and 805/81 combos is so miniscule that there's nothing detectable in ~99% of frames and only those particular special frames I was able to find can I even detect the tiniest bit of clipping difference! So really, don't worry about the exact DPLs for this phase of testing! ;)

SamuriHL

5th January 2021, 21:32

aron7awol

5th January 2021, 21:39

I don't have a great combo right now as I've been tweaking the crap out of it since these new builds have come out. Bouncing between a higher value (150) to lower values all the way down to 65. None of them really seem great in all content so far for me. But this was at 1000 dpl. Of course I've changed the dpl to 900 so that will almost definitely require going through content again. I was starting to lean towards somewhere around 75 ish the last time I messed around with it. God this is maddening. LOL
It doesn't really matter. Just pick one that seemed pretty good (75, 100, whatever) for now and then compare some equivalent combos to that. The exact DTN won't matter, the comparison will show the DPL differences only. Use this to hone in on DPL then worry about DTN later :)

SamuriHL

5th January 2021, 21:53

I see what you're saying. I was doing it backwards....yea that's a good idea. I'll see if I can find a few minutes tonight to at least give a cursory run through. The C8 should be good close to 900 DPL.

aron7awol

5th January 2021, 22:10

Here are the three formulas I use (always assuming "don't add peak nits" is checked):

DTN = multiplier * 50 / log10(DPL)
multiplier = log10(DPL) * DTN / 50
DPL = 10^(multiplier * 50 / DTN)

So if we pick 1000/75 as your "default", we get 913/76 and 836/77 as equivalent combos.

My guess is if you compare those 3 combos on the right specific frames, you'll perceive the tiniest clipping difference between 1000/75 and 913/76, and maybe the tiniest brightness difference between 913/76 and 836/77.

Edit: For anyone who wants to calculate these equivalent combos themselves, use the second equation to find the multiplier for your "default" DPL and DTN. Then increase/decrease DTN by 1 and then use the third equation to find the DPL for that DTN and the same multiplier you just found.

quietvoid

5th January 2021, 23:18

Okay, I had a quick look. Compared 1000/125 and 897/127 and got to the same conclusion. 1000 is clipping even though madVR has the same tonemap target for both.

https://i.ibb.co/q0njCBV/1000-125.jpg (https://ibb.co/q0njCBV) https://i.ibb.co/xggpVf2/897-127.jpg (https://ibb.co/xggpVf2)

Next is to see how high I can increase DPL before it becomes worse.

aron7awol

5th January 2021, 23:25

Okay, I had a quick look. Compared 1000/125 and 897/127 and got to the same conclusion. 1000 is clipping even though madVR has the same tonemap target for both.

Next is to see how high I can increase DPL before it becomes worse.
Well this is good, and expected, but at least now the only difference you should be seeing is that clipping and not an overall brightness change of the image (other than where it is clipping, which may be quite a bit in that particular one) due to a different curve being used.

I think the next logical steps would be to compare the 897/127 to 851/128 to see if there is another perceivable reduction in clipping, and also compare it to 947/126 to see if there is a perceivable increase in clipping.

quietvoid

6th January 2021, 00:19

Thanks. I did what you suggested and 851/128 brings a little more detail, while barely dimming.
947/126 is clipping just like 1000/125, and there's no noticeable improvement from 807/129.

Since I've switched to 100 DTN, I'll probably settle for 841/101.
No clipping and there's not much dimming other than those extreme FALL scenes (really, the horses one is the worst).
It's minimal and would rarely happen in real content.

For those who want to keep as much contrast, 851/128 would be it.
I think it makes The Meg skies a bit too dim and they look unnaturally saturated then, but anyone's free to want this.

I've looked at the clip you shared from BvS about saturation, I still prefer using desat 4 on top of desat 1, otherwise the lightning around the giant is really orange and looks dim.
Though I've lowered it to 100/100. Didn't try those low brightness shots with faces, but I'd assume those are no problem for us.

These are gonna be my (hopefully) final settings until madshi changes everything again: https://ibb.co/2YMt0tD
At least we've nailed DPL/DTN :)

aron7awol

6th January 2021, 01:04

Okay, so you found an improvement from 897 to 851, but not from 851 to 807. So your "optimal" DPL is probably somewhere between 851 and 897. Now if you want to go back to ~100 DTN, don't worry about using one of those exact combos, like 841/101. You can just use 850/100 or 875/100 or whatever you want, the difference is so small that now that we aren't directly comparing for clipping it doesn't matter. No reason to leave those 10 nits on the table :) Then you can go back to experimenting with just DTN now that you figured out your approximate optimal clipping point.

quietvoid

6th January 2021, 01:21

That's true, I'll have to see if it's worthwhile.

quietvoid

6th January 2021, 02:35

Well I compared 850 up to 880, there's no noticeable brightness increase.
Clipping starts to be noticeable at 880 DPL, but even then it's not something you'd notice when actually watching.
So it doesn't hurt me to set it at 870.

SamuriHL

6th January 2021, 02:46

That seems a reasonable DPL for these panels.

SamuriHL

6th January 2021, 04:16

Yea I'm liking that a LOT. Like a WHOLE lot. That fixes my crocodile issues in SM. The green spear looks fantastic. I've got DTN set to 127 at the moment and will play around with that more but 870 for the C8 is a very good DPL. Max lum, highlight sat I adopted your infinite (really unsure on this one), apply 2-4 100/100 is what I've been using and see no reason to change it. I did up my repair range from 0.010 to 0.020 but have no idea if that really matters. ceiling multiplier I still like at 4x. Yea, this is very good!

quietvoid

6th January 2021, 04:58

Repair range doesn't seem to do much (or anything) with max, it's noticeable with sep though.

SamuriHL

6th January 2021, 05:46

There's no reason for us to use sep that I've seen. I think we're pretty well dialed in. Obviously a little tweaking here and there but big questions have been answered.

Sent from my SM-G975U using Tapatalk

aron7awol

6th January 2021, 06:07

I think sep (or anything other than max) causes artifacts when used in combination with the non-defeatable TMing on the displays. At least it does for me on the green spear. I think maybe the two curves interacting end up causing a kink in the resulting curve or something. So it's max only for me, unless someone can point to an actual artifact caused by it. Every "artifact" I've seen has just been an accurate representation of the source.

SamuriHL

6th January 2021, 06:17

Very likely. As long as we can keep the options we have now we're in good shape.

Sent from my SM-G975U using Tapatalk

aron7awol

6th January 2021, 06:29

Yeah, I'm pretty happy with my current settings, too. The tunable ceiling makes a huge difference on The Meg. I still can't help but feel like the DTN/ceiling logic can still be optimized quite a bit more, but I actually think it would take someone taking the time to compile a whole spreadsheet of many different frames - FALL, FALL stddev, avgHL, and desired target (through experimentation on each frame) and then we'd have a shot at creating an algo to hit that whole range as closely as possible, but we might discover at that point we can do even better with some additional frame data. It's pretty good as-is, but I still think the target isn't quite where we want it in a lot of cases. Just comparing The Meg and S&M 10,000 there's a huge difference in the DTN and ceiling combo I like on each.

I'd actually be really interested in seeing what a neural network would come up with if fed all the different grades of the S&M demo with the goal of trying to create each of the lower grades from each of the higher grades using madVR :) Unfortunately it's one of the only things we have in so many different grades of the same exact content, but it does at least have a decent mix of scenes.

SamuriHL

6th January 2021, 16:53

aron7awol

6th January 2021, 17:41

That would be an awesome idea. I suspect they're going to focus on the PJ stuff a lot more from here but at least we got our wish for a few things that helped us a lot. I would argue that he should change the multiplier (since he only wants 1) to 4x or maybe even 5x. We'll have to tell him soon what we want there cause he doesn't want to leave the selectable option in place.
I feel very, very strongly that a hard-coded ceiling will not work. IMO, it works well at 4x on 10,000 S&M, but it's so much better at 3x for The Meg. I think the closest we can get to a hard-coded value will be for it to be some percentage of the DTN FALL multiplier, but I feel very strongly that won't work for everyone either.

There is absolutely just as much personal preference in the ceiling as there is with DTN in general. Please don't propose a hard-coded value!

SamuriHL

6th January 2021, 17:59

I wasn't proposing it. See madshi's response to me when he added it. He said he'd leave the selection there for a couple weeks but we had to decide what it needs to be. So this is problematic because he's expecting an answer on what the hard-coded value should be. So we may have to discuss that with him.

chros

6th January 2021, 18:25

Don't worry about this, he wants to take a look at DTN again along with different curves after desaturation, and that will take a long time! :D
Aron is right, just as with DTN value, there's no 1 fit all.
(I'll take a look at the samples today.)

SamuriHL

6th January 2021, 19:16

I wasn't arguing. :) I am simply repeating what madshi said to us when he added it. I tend to agree once size doesn't fit all but that's also problematic.

aron7awol

6th January 2021, 20:48

Oh, I know, I just think we need to try to convince him with the results of all of our experimentation that we can't even decide on a consistent multiplier from source to source, never mind user to user! :) I hoped that wouldn't be the case, but I realized that turned out to be wishful thinking after not very much testing.

Like I said, I liked 4x on S&M 10,000, I liked 3x on The Meg, and when I tested at 100 nit projector levels, I liked 2x or lower!

chros

7th January 2021, 00:24

SamuriHL

7th January 2021, 00:36

I've personally settled on quietvoid's 870 and think it's very well dialed in. Yes, DTN needs to be dealt with. And I agree, this is definitely improving our experience!!

aron7awol

7th January 2021, 01:00

Thank you guys for the samples, I tested all of them. Result: I don't see any (?) difference :D
(Tested for 2 hours, sitting at 1 meter, I think that tells a lot.)
So my DPL remain the same 700 and during the night 800.
Btw, aron, what DPL setting did you end up with?
And thanks for the fresh ideas, You managed to inject fresh blood into the circulation :D

So, if no other question left about DPL, can we move onto the next DTN algo topic?
Really strange! That seems impossible, but I believe you. :)

I'm not 100% decided, but that's mostly because the difference is so tiny! Something between 800 and 900. 875/80 is currently set as my default, but I can't promise I won't change that tomorrow. ;)

Sure, I'm ready to move on to the next topic.

quietvoid

7th January 2021, 02:40

Played with 119 a bit, not sure I'd use the new source nits clip option.
Clips very high brightness and therefore introduces hue shift in those very bright colors.
Though I can't see anything wrong with real content, it only seems to fix MMFR.

Also increased my DTN to 110.
I don't know what you've tested with The Meg @aron7awol but I can't see any difference between 3x and 4x with skies, the target stays the same.

And the new brightness variable desat is probably useless for us, most of the options are under our DPL.

SamuriHL

7th January 2021, 03:00

Again, at least in my opinion, I don't think we REALLY want to "fix" MMFR. I want to see the source material as they graded it. However that looks.

chros

7th January 2021, 12:04

Really strange! That seems impossible, but I believe you. :)

I'm not 100% decided, but that's mostly because the difference is so tiny! Something between 800 and 900. 875/80 is currently set as my default, but I can't promise I won't change that tomorrow. ;)
:D You see, it's not that easy at all.

Thanks for the formulas, I created other cases for me, e.g. with DTN 150 to not have big gaps between the resulting DPLs.
- with SM horses: I could only see slight differences in the snow, not much diff (if any) with 670-700 DPL, the interesting bit was that 800 DPL resulted in more details in the snow!!! :D But this scene has 1600 FALL!
- with SM crocodile: couldn't see any diff at all
- with bvs giant: not much diff again, maybe (?) 700 was better than 800
- with green spear: same as above

But here, as I said, 700 is a bit brighter than 800 because of the used LG tonecurve.

Again, at least in my opinion, I don't think we REALLY want to "fix" MMFR. I want to see the source material as they graded it. However that looks.
:) Well, that's you, but I can't watch that crap like that :)

Played with 119 a bit, not sure I'd use the new source nits clip option.
Clips very high brightness and therefore introduces hue shift in those very bright colors.
Though I can't see anything wrong with real content, it only seems to fix MMFR.
I quickly tried out this clip option: take a look at the end of the sample of mmfr, sundown, no change with 4000, but with 1000!!! This is how that should look like! (max lum, no other desat is used)

Sure, I'm ready to move on to the next topic.
Actually, I have 2 more related questions:
- what tone curve does madvr use with hdr output? is it the bt2048 or something else?
- don't we have patterns like this (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60393105) to determine easily DPL? (Vincent got 800 nits with his CX)
-- I guess we would need different maxCLL patterns for madvr (e.g. from 650-1000 in 10 step)

chros

7th January 2021, 12:13

OFF:

@aron7awol: if you use mpc-be/hc only (like I do, I don't use any frontend like Kodi) then checkout this android remote (https://forum.doom9.org/showthread.php?t=175007) for it and I made this small util for madvr (https://github.com/chros73/madvr-js-remote) (again using mpc-be/hc) with which you can easily switch between profiles. (@SirMaster has a standalone util (https://github.com/nicko88/HTPCRemote) that can be used with anything.)

@quietvoid: since you created a HDR10+ parser, do you know a way to remove HDR10+ data from video streams (not DoVi)? (maybe yusesope has it in his tool over makemkv forum, but I never used it). I'm asking because of this (https://forum.doom9.org/showthread.php?p=1919624#post1919624).
Thanks

QBhd

7th January 2021, 13:13

The Mehanik HDR patterns have all kinds of clipping patterns... they are free to download

QB

chros

7th January 2021, 13:50

Which ones exactly do you think will work for us?

aron7awol

7th January 2021, 16:56

- what tone curve does madvr use with hdr output? is it the bt2048 or something else?
- don't we have patterns like this (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60393105) to determine easily DPL? (Vincent got 800 nits with his CX)
-- I guess we would need different maxCLL patterns for madvr (e.g. from 650-1000 in 10 step)
It certainly uses the ST 2084 EOTF for HDR output since that's what the displays are expecting, and as for a curve, it should use the same BT.2390 curve whether it is outputting to SDR or HDR. For SDR output it has to translate to gamma.

I would say using the patterns are valuable, but they will really just tell us the clipping point of the TMing of the display rather than the actual DPL and/or how well it is tracking the EOTF to that point (your meter testing was much better in these aspects). I preferred the real content DPL/DTN combo approach only because it accounts for everything happening (clipping point, actual DPL, changed metadata) and shows what the actual result is on-screen for comparison purposes, which ultimately is what we care about anyway. If we can override metadata we can also eliminate that variable and choose the metadata that gives us the display TM we like best.

875 certainly clips more than 805 for me, but if it looks better to me, that's all that matters. That's really why I'm having so much trouble deciding, some clipping can be nice! I still love 1000 DPL in a lot of frames for this reason.

QBhd

7th January 2021, 18:05

I have always been happy with a little clipping... in order to eliminate clipping we then need to limit how bright the specular highlights can be, and that seems to defeat the purpose of HDR in my books. I am also in the camp of not fixing bad mastering. I am also very happy if most of the settings I have dialed in actually do nothing for the vast majority of the time. I think I have 5 betas downloaded and not a single one has been tested of yet, I am very happy with the 1000/65, compr limit disabled and avgHL ceiling disabled. When I get motivated I may look at some of the Mehanik clipping patterns to see if they can be useful, but they need to be small enough and granular enough to help us, otherwise we are just going to see a mixed bag of Pixel Shader and LG tone mapping all on top of one another.... Don't forget, with a C8 we can't turn off TM like you can with a CX.

QB

SamuriHL

7th January 2021, 19:21

I can tell you that the latest builds DO make a difference with our C8. Quite a bit, in fact. I haven't touched clipping yet but before that was added we dialed in a sweet spot so to speak for DPL. Try 870 and compare it to 1000 on scenes you like. You may like the difference and you won't lose brightness in the highlights.

aron7awol

7th January 2021, 19:35

I have always been happy with a little clipping... in order to eliminate clipping we then need to limit how bright the specular highlights can be, and that seems to defeat the purpose of HDR in my books.
Well, if we go exactly to the edge of clipping we shouldn't be limiting how bright the peaks of the highlights will be, other than technically the small range that the LGs use after the "clipping point", e.g. in @chros measurements, ~680-700. Not sure that 20 nits would be a really perceivable difference in practice, but in any case, clipping still ends up making highlights perceived as "brighter" overall, just because more pixels that make up the highlight end up clipped at max rather than being detail below max, and I agree that is often still an overall beneficial tradeoff as far as overall experience.

When I get motivated I may look at some of the Mehanik clipping patterns to see if they can be useful, but they need to be small enough and granular enough to help us, otherwise we are just going to see a mixed bag of Pixel Shader and LG tone mapping all on top of one another.... Don't forget, with a C8 we can't turn off TM like you can with a CX.
Well, we are currently always seeing a mixed bag of madVR and LG TMing combined. Which was kind of my point in just comparing the real content at the different DPL/DTN combos...I don't think we really care what exactly the interaction is of everything as much as the overall end result. As an example, if we could override metadata, @chros could run 960 metadata but run 680 DPL and get close to a pseudo-HGiG mode on his display, great tracking of the EOTF and no clipping. But more importantly, does he actually like the way that looks more than if he just runs 800 DPL or 900 DPL and sees the combination of both systems TMing and clipping to some extent?

I know my CX has HGiG mode but I haven't even messed with it yet, mostly because I saw reports that the measured peaks were significantly lower using that mode. But I will test it soon to see for myself.

QBhd

7th January 2021, 20:16

I don't have any scenes i like. What I look for is harm to any scene. It still makes no sense to me to limit the nits to the panel to anything below 900. It also makes zero sense to think that madVR will push nits to the panel that are above the DPL. 1000 has always been a guess, a guess that it will be enough nits for the tiniest of bright spots without clipping too much... Plus it also gives enough room for a gentle compression when it is needed.

QB

aron7awol

7th January 2021, 20:26

I don't have any scenes i like. What I look for is harm to any scene.
Okay, but how do you define "harm"? "Too much" clipping, or "too much" highlight detail lost? Or just something as simple as "looks worse"? That's all personal preference, or "liking" something more than something else, no?

It still makes no sense to me to limit the nits to the panel to anything below 900.
Why? Because the panel can actually hit 900 nits, or for some other reason?

It also makes zero sense to think that madVR will push nits to the panel that are above the DPL.
Who thinks this?

quietvoid

7th January 2021, 20:38

Another day, another test build. I compared max2 up to max4.
For MMFR, max3 completely removes the hue in flames/lightning, max2 just brightens it slightly.

In my Spider-Man Far from Home sample however, it seems to increase the highlight brightness but the TV doesn't like it and clips slightly around the lightning bolt, not much detail loss though.
I wouldn't use anything above max2 to avoid clipping, but this might just be an issue since we're double tonemapping..

In the saturation test pattern, all above regular max become more saturated somehow.
No idea what that translates to, I can't see differences in normal content.

Anyways, sticking to good ol' max.

QBhd

7th January 2021, 21:51

aron7awol

7th January 2021, 21:57

I would rank "harm" (I know terrible choice of words) in order as:
1) Lost detail
2) Dimming of image
3) Changing intent of creator
4) Artifacts
5) Clipping

Passthrough will be heavy in 1&5, too much DTN will be heavy in 2, etc... I am just looking for a balance that has as little of the above as possible

Yeah, 1 and 5 go hand in hand. 1000 DPL certainly has some of 1, as does any DPL over the LGs "clipping point" for the given metadata curve.

quietvoid

7th January 2021, 22:38

@quietvoid: since you created a HDR10+ parser, do you know a way to remove HDR10+ data from video streams (not DoVi)? (maybe yusesope has it in his tool over makemkv forum, but I never used it). I'm asking because of this (https://forum.doom9.org/showthread.php?p=1919624#post1919624).
Thanks
https://forum.doom9.org/showthread.php?p=1932777#post1932777

aron7awol

8th January 2021, 01:21

Re: EOTF tracking and rolloff/clipping points, is there a reason you guys don't upload DeviceControl templates with rolloff point of 100% to stop the TV from TMing and instead tell it to hard clip at peak luminance?

I can't say I've done it myself yet, but it's something I've been considering doing after I measure the actual peak luminance of my display.

quietvoid

8th January 2021, 01:35

I think that feature is only available to 2019+ models :D

aron7awol

8th January 2021, 01:40

Oh, sorry, I was thrown off because it showed 2017 and 2018 support for DeviceControl, but now I see that was not for the PQ Curve templates. :(

quietvoid

8th January 2021, 01:41

Not for the PQ curve upload.

SamuriHL

8th January 2021, 04:07

Yea, screw you C9 + owners! :P We'd LOVE that option but it's not available to us. :(

aron7awol

8th January 2021, 05:09

So madVR has the feature to detect hard coded black bars, and one of the options when it does is to shift the image to the bottom of the screen. I'm trying to accomplish the same thing but on videos that don't have hard coded black bars. Essentially I want to put both black bars on the top of the screen and slide the image to the bottom. Are you guys aware of any way to get madVR to do this or any player that can do this?

Asmodian

8th January 2021, 06:20

QBhd

8th January 2021, 07:14

Okay, so I was motivated to see what the Mehanik Clipping Patterns offered for us. And the results were odd...

So these are the two files that I mainly focused on:

http://qbstorrents.erebus.feralhosting.com/Doom9/madVR.Tonemapping/02.%20White_Color%20clipping.rar

I disabled all the settings that I could in the Pixel Shaders and this was just a straight up DPL experiment.

Now for the oddness...

If I lowered DPL, more of the boxes were visible I even went down to 400 for shits and giggles, but even at 800 more of the 4000 nit file was visible. It was noticeably dimmer, especially when toggling to Passthrough. This seems to make sense... Then I went above 1000 DPL, again even as high as 2000... Instead of less boxes being visible and clipping kicking in, more boxes were again visible. More to the point, the clipping mark was roughly the same at 800 DPL as at 1200 DPL. So I dove deeper, 950 & 1050 both clipped roughly the same with more boxes being visible than 1000. So it seems that if you were to chart the last visible box vs the DPL it would be a "U" shaped chart with 1000 DPL the low point (roughly speaking).

So what does this all mean? Is it even more proof that 1000 DPL on the C8 is optimal? What is causing the clipping to shift UP when DPL goes above 1000? I know the TV's TM'ing will play a roll, but not this much...

Thoughts?

QB

Edit: oops links not doing what I wanted... I will fix now
Edit2: Fixed (I hope, LOL)

aron7awol

8th January 2021, 15:48

^^^ Those are the same exact two patterns that I use for this stuff, so I can speak to my observations on them as well.

I believe the reason you are seeing what you are seeing is that once you go >1000 DPL, and in turn >1000 maxCLL, you are seeing the result of the LG TMing changing from its 1000 nit curve to something in between that and its 4000 nit curve. We know on the 2019/2020 models it interpolates between the curves when maxCLL falls in between, and I would assume it does the same on earlier models even though we can't use the templates to modify the curves on those models.

So I don't think it means 1000 DPL is optimal, necessarily. I think you're saying it clips the most at 1000, right? If not losing highlight detail is truly your #1 priority, I'd say something <1000 is "optimal".

But the real issue is we need to be able to modify the metadata in order to not to be a slave to the display's TMing changing with DPL. Without that, 1000 DPL is just what results in the most clipping (and loss of highlight detail).

aron7awol

8th January 2021, 15:49

In processing -> zoom control there is a setting "always shift the image" that can be set to the top or bottom of the screen. This works even when there are no black bars.

If you actually want to avoid using a portion of your screen entirely, you can set the display type to a projector to access the "screen config" menu and define a visible screen area.
Thanks. I know I tried that in the past (with the old Kodi DSPlayer only) and it didn't work, and maybe the player was why, so I'll certainly test it again!

Edit: Sure enough, it works in MPC. Thanks!

QBhd

8th January 2021, 16:48

As someone with an engineering background, when I see a clear inflection point it means something. And yes, until we are not slaves to DPL linked to metadata I will stick with what the data is trying to show us

Also, there is still less clipping than with Passthrough... And detail can also be retained with DTN (this test was just to find the DPL)

QB

aron7awol

8th January 2021, 16:59

As someone with an engineering background, when I see a clear inflection point it means something. And yes, until we are not slaves to DPL linked to metadata I will stick with what the data is trying to show us
It certainly does mean something, it's the point beyond which it starts changing to a totally different curve. You see more and more clipping as you set DPL beyond the display's capabilities, up to 1000 DPL, and then beyond that point the TV starts switching toward the 4000 nit tone curve and clipping is reduced.

If you want the most clipping possible from the TV without being able to override metadata, 1000 DPL gives you that. I was just pointing out that "the most clipping possible" flies in the face of your priority list, with #1 being not losing detail. But if you prefer the way 1000 looks with real content, perhaps your priority list is just not what you thought it was. ;)

Also, there is still less clipping than with Passthrough...
Yes, this is true, but there is still more clipping with 1000 DPL than a lower DPL in your testing, right?

And detail can also be retained with DTN (this test was just to find the DPL)
More detail can be retained with DTN, yes. And yet, still, if you compare 1000 DPL and a lower DPL with the same target, you will still find there is even more detail (due to less/no clipping) with the lower DPL. This will always be the case, because if DPL is set to something exceeding the actual peak nits of the display, madVR will send pixels that exceed and will be clipped by the display.

Again, I'm not trying to tell you what you prefer and what you don't prefer, just pointing out differences. If you prefer the overall compromise of 1000 DPL best, with its balance of some clipping and overall picture brightness, I can totally relate. I really like the way it looks as well.

QBhd

8th January 2021, 18:16

aron7awol

8th January 2021, 20:11

You're preaching to the choir. More detail (i.e. less clipping with this particular test) can be attained in two ways. Lowering peak brightness or increasing peak brightness... All I am saying is the balance lies in the middle and this test is very easy to find that DPL +/- 50 nits. 1000 is a nice round number which makes some sense all by itself... And the flashing box at 860 (on the pattern) is so incredibly faint for both 1050 & 950 (DPL) that narrowing it down further would be difficult.
I'm just not sure what you are actually trying to balance or narrow down between 1050 and 950?

If you're trying to find the point with the most clipping, that's 1000. But I thought you didn't want to clip and lose detail, which is why I find it strange that you're honing in on the point that clips the most. And really, there's nothing to hone in on anyway, we know that's 1000.

chros

8th January 2021, 23:13

OFF:
https://forum.doom9.org/showthread.php?p=1932777#post1932777
Thank You! It works like a charm! :) (You see there are people who don't understand what the problem is / you want to achieve and just complaining non-stop :))
What popped into my mind is - if you will have time / in the mood -:
- you could incorporate this into the hdr10+ parser project
-- by setting a flag that allows stripping the metadata while making a backup of it in one go
-- allowing to rewrite the original hdr10+ file with the saved metadata

ON:
And all this because bloody LG's magenta/cyan issue with HDR10 content!!! :D It just drives me nuts, right @jk82?! :D

Btw, @aron7awol, would you mind testing the cyan/magenta issue (https://forum.doom9.org/showthread.php?p=1925528#post1925528) on your CX as well? But beware, once you've seen this (if it exists), you won't be able to unsee it anymore! :D Thanks

chros

9th January 2021, 00:13

Are you guys aware of any way to get madVR to do this or any player that can do this?
What do you need this for?

see what the Mehanik Clipping Patterns offered for us. And the results were odd...

Now for the oddness...

Thoughts?

Thanks, something is indeed strange, the question is why.

1. here, both with the 1000 and 4000 file (both in nomarl and PC mode):
- a bit more visible the last box with 800 DPL than as with 700DPL! (Remember, that's what I saw with the SM horses scene in the snow)
- at 650 and 840 (!) boxes starts to disappear at the high end
- between 700 and 820 all the boxes appear, although the last box is pretty faint with 700

So, what does this mean? Is it because of the double tonemapping?!
It seems that the optimal settings for me are between 740-820, but that is way higher (40-120 nits) that I measured as a 1% window peak.

2. maybe it's not important for us, but I think for determining the peak DPL these files are not good, because they result in double tonemapping (unless you use 100 and 4000 DPL for the corresponding files).
- that's what I meant when I mentioned Vincent's video, that we would need the right maxCLL file with the set DPL in madvr, e.g.:
-- 800 maxCLL file for 800 DPL, 850 maxCLL file for 850 DPL, etc

aron7awol

9th January 2021, 06:13

Btw, @aron7awol, would you mind testing the cyan/magenta issue (https://forum.doom9.org/showthread.php?p=1925528#post1925528) on your CX as well? But beware, once you've seen this (if it exists), you won't be able to unsee it anymore! :D Thanks
You're making me afraid to test this. The last thing I need is another thing to be looking out for when I'm trying to watch stuff! I'm bad enough already.

aron7awol

9th January 2021, 06:33

What do you need this for?
I wanted to move non-16:9 content to the bottom of the screen. It turns out it does work on the old Kodi DSPlayer as well as MPC. I tested it tonight and I actually love it. It not only moves the video closer to eye level, but it also moves the video closer to the center channel for better anchoring of the dialogue. Big upgrade! It's the little things :)

Thanks, something is indeed strange, the question is why.

1. here, both with the 1000 and 4000 file (both in nomarl and PC mode):
- a bit more visible the last box with 800 DPL than as with 700DPL! (Remember, that's what I saw with the SM horses scene in the snow)
- at 650 and 840 (!) boxes starts to disappear at the high end
- between 700 and 820 all the boxes appear, although the last box is pretty faint with 700

So, what does this mean? Is it because of the double tonemapping?!
It seems that the optimal settings for me are between 740-820, but that is way higher (40-120 nits) that I measured as a 1% window peak.
Your display certainly has some strange behavior with different maxCLL values, as you found in your curve testing. You really need the metadata override!

2. maybe it's not important for us, but I think for determining the peak DPL these files are not good, because they result in double tonemapping (unless you use 100 and 4000 DPL for the corresponding files).
I agree.

- that's what I meant when I mentioned Vincent's video, that we would need the right maxCLL file with the set DPL in madvr, e.g.:
-- 800 maxCLL file for 800 DPL, 850 maxCLL file for 850 DPL, etc
But doesn't madVR already change maxCLL to match DPL, meaning that is already happening for you?

chros

9th January 2021, 13:14

You're making me afraid to test this. The last thing I need is another thing to be looking out for when I'm trying to watch stuff! I'm bad enough already.
:D I bet your curiosity will win in the end :) (It's enough to check The Witcher sample against its DV version on Netfilx.)

I wanted to move non-16:9 content to the bottom of the screen. It turns out it does work on the old Kodi DSPlayer as well as MPC. I tested it tonight and I actually love it. It not only moves the video closer to eye level, but it also moves the video closer to the center channel for better anchoring of the dialogue. Big upgrade! It's the little things :)
Interesting.

Your display certainly has some strange behavior with different maxCLL values, as you found in your curve testing. You really need the metadata override!
I agree, let's ask madshi all of us at the same time!
But I think that's what @QBhd saw as well on his C8, isn't it?

But doesn't madVR already change maxCLL to match DPL, meaning that is already happening for you?
It does change it indeed, but my point was that it also does tonemapping with these 2 files (e.g. with 900 DPL and with the 1000 maxCLL file), the optimal would be if we would have maxCLL files for all the DPLs we want to try out; that's why we have double tonemapping with these.
Although you could argue that this is exactly what's happening with real content as well...

aron7awol

9th January 2021, 23:05

It does change it indeed, but my point was that it also does tonemapping with these 2 files (e.g. with 900 DPL and with the 1000 maxCLL file), the optimal would be if we would have maxCLL files for all the DPLs we want to try out; that's why we have double tonemapping with these.
Although you could argue that this is exactly what's happening with real content as well...
But with how madVR has worked to this point, do you really have 900 DPL and 1000 maxCLL combo? Because madVR changes maxCLL to 900 in that 900 DPL scenario anyway. So really, you already have 900/900 combo.

SamuriHL

9th January 2021, 23:59

That's also why I asked for madshi to allow metadata override for passthrough. We want to be able to see how metadata alone messes with how the TV itself works. Now, yes, I could already do that myself, but I don't feel I'm fully qualified to judge what's happening and more eyes are better.

chros

10th January 2021, 19:39

But with how madVR has worked to this point, do you really have 900 DPL and 1000 maxCLL combo? Because madVR changes maxCLL to 900 in that 900 DPL scenario anyway. So really, you already have 900/900 combo.
Yes, that's true, but my point was that if the file indeed has data up to 1000 nits then it means that madvr has to compress it to 900!
But if we would have a 900 maxCLL file madvr (in theory) wouldn't compress anything with 900 DPL :)

Anyway, now I think we can move on to DTN algo (since everybody has their own DPL setting)...

What do you think?
- first we have to come up with a DTN value for "normal" sources:
-- these are above of our DPL but not much
-- they only have "reasonable" FALL values

I think The Grinch (2018) (CLL can reach 1300-1600, FALL 200-300) is a good candidate for this first step, if everyone agrees. (If you know other titles like this then let us know.)

So, I created a google-sheet (https://docs.google.com/spreadsheets/d/1IupvatPOHJxpzEiH_8JqHO5QJkKBRNhEUUpt3m0_hNI/edit?usp=sharing) for all of us (let me know if you miss something):
- used formula: APDL = log10(DPL) * DTN/50 * FALL
- 2 tables in it for: 800 DPL and 875 DPL
-- first column is DTN, the first row is FALL

This doesn't take into account the followings (for now):
- CLL of the frame
- limiter

aron7awol

10th January 2021, 20:00

Yes, that's true, but my point was that if the file indeed has data up to 1000 nits then it means that madvr has to compress it to 900!
But if we would have a 900 maxCLL file madvr (in theory) wouldn't compress anything with 900 DPL :)
Okay, if that's what you're trying to accomplish, can't you set madVR to clip at 900 nits?

Edit to add: I still don't think it's possible to figure out to actual DPL using test patterns without being able to disable TMing on the display. Vincent is only able to do that because he knows/assumes the display is just tracking the EOTF and then clipping at DPL. It's not the double TMing that makes it's difficult/impossible, it's any TMing at all. So even if we could disable the display's TMing, we wouldn't use madVR TMing either for this exercise.

chros

10th January 2021, 20:15

Okay, if that's what you're trying to accomplish, can't you set madVR to clip at 900 nits?
Maybe I can now, good idea, but previously we couldn't :)

About DTN, let's assume the limiter doesn't have any effect (only couple of really bright highlights are in the frame):
- if we take a look at the 875 table, we see that even using 60 DTN, DTN will kick in with 250 FALL!
- now the question is how high the CLL?
-- because if it's only 850 (below DPL) then we don't want DTN to work at all for sure! (I'm not sure that's the case now)
-- even if it's let's say 1100/1300 CLL, then the question is whether it's necessary to work or not

aron7awol

10th January 2021, 20:29

About DTN, let's assume the limiter doesn't have any effect (only couple of really bright highlights are in the frame):
- if we take a look at the 875 table, we see that even using 60 DTN, DTN will kick in with 250 FALL!
- now the question is how high the CLL?
-- because if it's only 850 (below DPL) then we don't want DTN to work at all for sure! (I'm not sure that's the case now)
-- even if it's let's say 1100/1300 CLL, then the question is whether it's necessary to work or not
We can't really assume this frame only has a couple really bright highlights if it has a 250 FALL. 250 FALL is really quite high. DTN should almost certainly be kicking in on 250 FALL frames. (This is why on my Excel sheet that I posted some previous DTN calculation screenshots from, you'll see I didn't use a linear scale for FALL and instead started really low and did exponential growth. I'll post what I used.)

If frame peak (I think that's what you are talking about when you say CLL, right?) is <DPL DTN will do nothing. In reality, DTN will do nothing in most cases even when frame peak > DPL, because that happens all the time at FALL values that are too low for DTN to kick in.

Edit: I started with FALL of 2 and multiplied by 1.3 each time, which resulted in FALL >4000 within the ~30 rows.

Edit2: I also made the formula ADPL = max(log10(DPL) * DTN/50 * FALL, DPL) so that ADPL = DPL when DPL is greater. Then we see it TMs to DPL until it exceeds it and kicks in.

The example I posted for Neo-XP:

https://i.imgur.com/ounJX9g.png

chros

10th January 2021, 21:55

Yes, CLL is frame peak, Edit2 is good idea, Fall values also can be exponential, yes.
But we need numbers for us, that's why I crsated the sheet.
And I'm not sure at all that 250 Fall has to trigger DTN with 60 DTN in every cases.

aron7awol

10th January 2021, 22:32

I think I'd kind of summarize my interpretation of the whole operation of DTN and the avgHL ceiling as follows:

1. Frame peak is pretty much irrelevant. DTN doesn't even look at it, and it can be way higher than DPL without the highlights driving FALL high enough for DTN to kick in. Most frames including those with really bright highlights fall into this category, and DTN does nothing. madVR TMs to DPL.
2. Now as we get into frames with higher FALL (let's just say >200 FALL for this discussion, but obviously it depends on DPL and DTN settings) DTN kicks in and madVR TMs to a target higher than DPL then scales the result down to peak at DPL. Essentially (with don't add peak nits checked) this target is just a simple multiplier of FALL which we can calculate using the formula I posted previously.
3. Within this subset of frames with higher FALL, there is an additional consideration. Is this whole frame "bright" (I believe currently hardcoded to >100 nits but would love to be able to change this) or does it have dark/normal areas and also quite big and bright areas? This is where the avgHL ceiling comes in. avgHL is always >= FALL, but on frames where pretty much the whole frame is bright, it will be closer to FALL, and on frames where there are normal/dark areas, it will be significantly higher than FALL. I think essentially we are trying to account for frame dynamic range differences, and this helps us, to a limited extent, to differentiate within these high-FALL frames between those that have more or less dynamic range. And those with less dynamic range probably don't need as high of a target, and since avgHL will be close to FALL in these cases, the lower ceiling multiplier will result in a lower target than the normal target arrived at via the higher FALL multiplier. So we can try to choose a FALL multiplier that results in proper targets for the higher dynamic range high-FALL frames, but also an avgHL multiplier that results in proper (lower) targets for the lower dynamic range high-FALL frames.

I hope that is all coherent! I wanted to post something similar to this in the AVS thread for discussion there, but I've been waiting for madshi to get back on this topic. Meanwhile, we can work on it here.

Anyway, I'm not sure avgHL ceiling is the optimal way to account for category 3 and these dynamic range differences (and this is why I was thinking about things like stddev/percentile as potentially much better ways), but having the 100 nits "highlight" definition adjustable might help. I actually think the best way to hone in the optimal formulas would be to do a sort of case study where we pick a number of higher-FALL frames from a movie or different movies that sort of fall into these different scenarios, and each of us goes to each of these frames and finds their preferred target. Then based on this result we should be able to come up with a combination of settings and/or an improved formula that gets us close to these targets automatically in each case.

aron7awol

10th January 2021, 22:44

And I'm not sure at all that 250 Fall has to trigger DTN with 60 DTN in every cases.
This obviously depends on the specified DPL/DTN combo, but I'm confident the target exceeds DPL with FALL of 250 on the vast majority of combos that people run.

Low DPLs will exceed it pretty much automatically, as will

700/50
750/53
800/56
850/59
900/61
950/64
1000/67

I think all of us are running a combo that has a higher DTN that those.

You're running 700/60 and 800/?
I'm running a wide variety of DPLs and ~80 DTN
IIRC the other guys here are all ~75 DTN or higher?

Don't forget our DTNs all increased (or should have) with the "don't add peak nits" option, which is where this formula applies. So I don't think any of us are running low (<60) DTNs anymore with that option. You might be the lowest at 700/60.

chros

11th January 2021, 16:13

Fall values also can be exponential, yes
But I created those only up until 400 because we want to test "normal" content first, right?

This obviously depends on the specified DPL/DTN combo, but I'm confident the target exceeds DPL with FALL of 250 on the vast majority of combos that people run. ...
...
You're running 700/60 and 800/?
...
Don't forget our DTNs all increased (or should have) with the "don't add peak nits" option, which is where this formula applies. So I don't think any of us are running low (<60) DTNs anymore with that option. You might be the lowest at 700/60.
I use 700/60, 800/59. And yes, obviously "don't add peak nits" option is enabled.
We want to experiment with the same settings, to avoid confusion, like this!

So, what about 800/75 to start with? If you like other pair instead, we can use that. But in this case we only have to take a look at 1 row in that sheet.
Along with these options:
- "don't add peak nits" enabled
- "avgHL ceiling" disabled (for "normal content" we don't need it)

I think I'd kind of summarize my interpretation of the whole operation of DTN and the avgHL ceiling as follows:

1. Frame peak is pretty much irrelevant. DTN doesn't even look at it, and it can be way higher than DPL without the highlights driving FALL high enough for DTN to kick in. Most frames including those with really bright highlights fall into this category, and DTN does nothing. madVR TMs to DPL.
Yes, that's the case for now, but the question is whether this is allright? We agree that having 300 FALL with 600 frame peak is not the same as having it with 1300 frame peak.
That's why I though about CLL, but maybe your 3. point is better for this.

3. Within this subset of frames with higher FALL, there is an additional consideration. Is this whole frame "bright" (I believe currently hardcoded to >100 nits but would love to be able to change this)
I also think it's 100 nits, the brightness histogram (top right on the OSD) display these as grey (first half of the histogram, 0-63).
And as I said, probably it needs to be some dynamic detection for this, based on the content.

or does it have dark/normal areas and also quite big and bright areas? This is where the avgHL ceiling comes in. avgHL is always >= FALL, but on frames where pretty much the whole frame is bright, it will be closer to FALL, and on frames where there are normal/dark areas, it will be significantly higher than FALL. I think essentially we are trying to account for frame dynamic range differences, and this helps us, to a limited extent, to differentiate within these high-FALL frames between those that have more or less dynamic range. And those with less dynamic range probably don't need as high of a target, and since avgHL will be close to FALL in these cases, the lower ceiling multiplier will result in a lower target than the normal target arrived at via the higher FALL multiplier. So we can try to choose a FALL multiplier that results in proper targets for the higher dynamic range high-FALL frames, but also an avgHL multiplier that results in proper (lower) targets for the lower dynamic range high-FALL frames.
Good idea, the histogram can help with determining this (instead checking CLL).

Anyway, I'm not sure avgHL ceiling is the optimal way to account for category 3 and these dynamic range differences (and this is why I was thinking about things like stddev/percentile as potentially much better ways)
What do you mean about stddev? (There's a FALL Stddev on OSD as well, but no one knows what it shows :) )

but having the 100 nits "highlight" definition adjustable might help.
This has to be dynamic, we don't want to use profiles for this.

I actually think the best way to hone in the optimal formulas would be to do a sort of case study where we pick a number of higher-FALL frames from a movie or different movies that sort of fall into these different scenarios, and each of us goes to each of these frames and finds their preferred target. Then based on this result we should be able to come up with a combination of settings and/or an improved formula that gets us close to these targets automatically in each case.
Hmm, that's one way, but I'm not sure it would be fruitful.
What about this for now?
- select 1 title/scenes with appropriate frames for testing
- we will check it together
- try to draw the same (!) conclusion (if we won't agree what we should achieve then we don't have to work on this together :) )

So, I propose (for the 3rd time) The Grinch (2018).
Thoughts?

aron7awol

11th January 2021, 20:54

But I created those only up until 400 because we want to test "normal" content first, right?
Sure, anything is fine with me.

I use 700/60, 800/59. And yes, obviously "don't add peak nits" option is enabled.
We want to experiment with the same settings, to avoid confusion, like this!

Okay, I didn't realize you were trying to get us all to agree on a DTN setting? Seems unlikely, but I'll go along. :)

So, what about 800/75 to start with? If you like other pair instead, we can use that. But in this case we only have to take a look at 1 row in that sheet.
Along with these options:
- "don't add peak nits" enabled
- "avgHL ceiling" disabled (for "normal content" we don't need it)

800/75 is fine. avgHL ceiling won't be needed if by "normal content" you mean higher FALL but never also high avgHL (overall very bright scenes). I don't know The Grinch well enough to say whether such scenes exist, but I'll take your word for it. :)

Yes, that's the case for now, but the question is whether this is allright? We agree that having 300 FALL with 600 frame peak is not the same as having it with 1300 frame peak.
That's why I though about CLL, but maybe your 3. point is better for this.
Yeah, I think we need significantly more information than just frame peak and FALL. Because there could easily be two frames with equal peak and FALL and yet one is overall very bright and could use a lower target (ceiling) than FALL (or peak) would otherwise suggest.

I also think it's 100 nits, the brightness histogram (top right on the OSD) display these as grey (first half of the histogram, 0-63).
And as I said, probably it needs to be some dynamic detection for this, based on the content.
Agreed.

Good idea, the histogram can help with determining this (instead checking CLL).

What do you mean about stddev? (There's a FALL Stddev on OSD as well, but no one knows what it shows :) )
Yes, I'm thinking some combination of a different from 100 avgHL value, stddev, percentile, histogram will give us the information we need to determine this far more accurately.

This has to be dynamic, we don't want to use profiles for this.
The main reason I want this adjustable is to try some different values on different scenes and see what the result is, to use this data as part of figuring out a dynamic formula. Because I can't calculate an avgHL > 200 nits, for example, myself, I need to be able to set it to 200 and see what it spits out.

Hmm, that's one way, but I'm not sure it would be fruitful.
The reason I suggested this approach is that it doesn't require us to agree on a DTN value, which there seems to be nothing even close to a consensus on amongst users now. It would give us a number of preferred targets for each user across a variety of types of frames, and since the goal is to come up with a new target algo that works for everyone and multiple DPL/DTN combos, we'd be building a formula that already handles this variety. Essentially, it's the relative differences in preferred targets across the different types of frames that we really need to create a formula to handle. Then we could do some sort of regression model to get from the information we have to the preferred targets. It would be really helpful to see if those relative differences are reasonably consistent from user to user even with different settings. I would imagine they would be. Essentially I'm wanting to sort of take a step back, forget how everything works right now, and each of us goes into each frame with nothing but a goal of just finding a specific TM target (in nits) they like best, and then with all that data, we figure out the best way to get there. It may be similar to how it works now, or it may be drastically different. It's like just letting the data guide us. But that being said...

What about this for now?
- select 1 title/scenes with appropriate frames for testing
- we will check it together
- try to draw the same (!) conclusion (if we won't agree what we should achieve then we don't have to work on this together :) )

So, I propose (for the 3rd time) The Grinch (2018).
Thoughts?
...I'm certainly willing to participate in a group testing of a single DPL/DTN combo if that's what everyone wants to do. The data will be useful in any case. And The Grinch is fine with me. So what are you going to pick a number of frames from that movie for each of us to test?

chros

11th January 2021, 23:50

The main reason I want this adjustable is to try some different values on different scenes and see what the result is, to use this data as part of figuring out a dynamic formula. Because I can't calculate an avgHL > 200 nits, for example, myself, I need to be able to set it to 200 and see what it spits out.
Agreed in this. Especially if we suppose that the nominal white of normal content today is 203 nits.

The reason I suggested this approach is that it doesn't require us to agree on a DTN value, which there seems to be nothing even close to a consensus on amongst users now.
I understand and we can use this later, but not now: too many variables mixed with personal preference without having common understanding of our goal, we will fail miserably :)
I think it's important to go through couple of titles (not just 1) at first with the same settings, to establish the goal. This alone will be complicated enough for now :)

It would give us a number of preferred targets for each user across a variety of types of frames, and since the goal is to come up with a new target algo that works for everyone and multiple DPL/DTN combos, we'd be building a formula that already handles this variety. Essentially, it's the relative differences in preferred targets across the different types of frames that we really need to create a formula to handle. Then we could do some sort of regression model to get from the information we have to the preferred targets. It would be really helpful to see if those relative differences are reasonably consistent from user to user even with different settings. I would imagine they would be. Essentially I'm wanting to sort of take a step back, forget how everything works right now, and each of us goes into each frame with nothing but a goal of just finding a specific TM target (in nits) they like best, and then with all that data, we figure out the best way to get there. It may be similar to how it works now, or it may be drastically different. It's like just letting the data guide us. But that being said...
That's a really good idea for the future!

...I'm certainly willing to participate in a group testing of a single DPL/DTN combo if that's what everyone wants to do. The data will be useful in any case. And The Grinch is fine with me. So what are you going to pick a number of frames from that movie for each of us to test?
You can pick up scenes as well, share the timestamps and that's it.
- let's use 800/75 then for now
-- "disabled avgHL ceiling" enabled
-- "don't add peak nits" enabled
-- all the other unnecessary options disabled (contrast/shadow recovery, dynamic clipping, sky strength)
- that means DTN will kick in at ~184 FALL (if frame peak > DPL)
- the question is what FALL/scene combo DTN is really needed and is there any scene in the title when DTN is unnecessary with these settings
-- if we can come up with a DTN value at the end for this title then that'd be the best :)

One more important thing:
- madvr can/does result in different ADPLs during playback vs still images!

The Grinch is very bright, so it's easy to find scenes like this:
- 1: 00:02:05 - 00:03:20 (2985 - 4784)
- 2: 00:06:45 - 00:07:25 (9699 - 10658)
- 3: 00:13:10 - 00:14:00 (18930 - 20130)
-- this scene is interesting, there are frames when frame peak is only ~430 but FALL is >200.
- 4: 00:16:00 - 00:16:45 (23006 - 24086)

aron7awol

12th January 2021, 01:44

I understand and we can use this later, but not now: too many variables mixed with personal preference without having common understanding of our goal, we will fail miserably :)
I think it's important to go through couple of titles (not just 1) at first with the same settings, to establish the goal. This alone will be complicated enough for now :)
Please don't just assume it would fail miserably. The personal preference is something I absolutely want to capture in the data, because the end goal is to create something that works for all the different personal preferences. After all, DTN is tunable for a reason, it's just not a one-size-fits-all setting. Some people prioritize a brighter image and some people prioritize depth and detail. There's no right or wrong value/approach. I just want to be clear that what I'm suggesting is that for a given frame, each of us doesn't say what DTN to use for that frame or what ceiling to use for that frame or anything complicated at all, but literally just the target they liked best (say, ~1450 nits for this particular frame, ~2560 nits for that frame, etc.). I think that if everyone did that and I compiled that data, I'd be willing and able to come up with an algo fairly quickly. It's fine though, I should still be able to get some data from your experiment for this purpose.

- the question is what FALL/scene combo DTN is really needed and is there any scene in the title when DTN is unnecessary with these settings
-- if we can come up with a DTN value at the end for this title then that'd be the best :)
I can't imagine how we can possibly have consensus on this based on the wide range of personal preference and DTNs people run right now, but I'm open-minded, we'll see! :)

One more important thing:
- madvr can/does result in different ADPLs during playback vs still images!
I think this is due to scene detection. Switching profiles while paused causes it to reset, so I think that's what we should do, for consistency. These dynamic changes due to scene detection are a necessary evil in practice, but IMO we shouldn't worry about or account for that in this exercise, it would muddy the waters and throw off the real desired targets.

aron7awol

12th January 2021, 05:09

Alright, I did a quick round of testing:

https://i.imgur.com/LFltauh.png

Notes:
1. I calculated avgHL by finding a ceiling that kicked in and dividing the resulting ceiling by the multiplier.
2. DPL/avgHL is intended to show what the ceiling would have to be in order to pull the target down to DPL (disable DTN on that frame essentially)
3. Switching profiles works well to reset scene detection.
4. I probably should have noted peak as well in case it can be helpful. Oh well, next time.
5. It would also be helpful to do this by frame number rather than timestamp so we can use the same exact frames and/or go back to the same frames for further research.
6. Sky algo is also an important aspect of this. Despite it interacting with the frame measurements, I feel like if there is consensus that sky detection is useful, we should leave it on since it is helpful in reducing the measurements in certain scenarios which only helps us here.

chros

12th January 2021, 11:26

I think this is due to scene detection. Switching profiles while paused causes it to reset, so I think that's what we should do, for consistency. These dynamic changes due to scene detection are a necessary evil in practice, but IMO we shouldn't worry about or account for that in this exercise, it would muddy the waters and throw off the real desired targets.
I guess so too, but I think it's important to watch each scene as a whole as well at the end because that's what happening during normal viewing :)
I know it's another variable in the equation but it's important to understand how the whole thing (can) work in "real life".
E.g. at the end of scene 4 (btw I added scene numbers to the post above) there is unnecessary dimmed image due to how this works.

1. I calculated avgHL by finding a ceiling that kicked in and dividing the resulting ceiling by the multiplier.
2. DPL/avgHL is intended to show what the ceiling would have to be in order to pull the target down to DPL (disable DTN on that frame essentially)
3. Switching profiles works well to reset scene detection.
4. I probably should have noted peak as well in case it can be helpful. Oh well, next time.
5. It would also be helpful to do this by frame number rather than timestamp so we can use the same exact frames and/or go back to the same frames for further research.
Thanks, looks interesting, especially the avgHL and DPL/avgHL!
We can also take a look at the histogram.
Good idea about frame peak and frame numbers.

One more thing popped into my mind: have you ever used madmeasure.exe? (It only requires the mkv as param, I'm not sure whether it works with BDMV directories.) It will create a ~100MB measurement file of the content, and madshi added tools (https://forum.doom9.org/showthread.php?t=146228) into the first post that can analyse its content (e.g. by frames), "Simple Analyzer for madMeasureHDR".

6. Sky algo is also an important aspect of this. Despite it interacting with the frame measurements, I feel like if there is consensus that sky detection is useful, we should leave it on since it is helpful in reducing the measurements in certain scenarios which only helps us here.
Let's forget about the sky algo (and the rest) for now until it's fixed, because it's buggy (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60341772) (meaning unusable), (and I think it's not needed with "normal" titles, but not use).

PS: I only have limited amount of time per day for this, but I'll try to do testing each day and think with you.

aron7awol

12th January 2021, 16:34

I guess so too, but I think it's important to watch each scene as a whole as well at the end because that's what happening during normal viewing :)
I know it's another variable in the equation but it's important to understand how the whole thing (can) work in "real life".
E.g. at the end of scene 4 (btw I added scene numbers to the post above) there is unnecessary dimmed image due to how this works.
Scene detection is part of normal viewing, but it's important to note that the reason we see this change in target is due to the previous content (and it trying to make smooth changes in TMing rather than abrupt changes), not the current content. So if we make a decision in targeting based on this, we're just changing our target based on previous content, but whether some previous content was significantly brighter or dimmer than the current content is essentially "random", and so it has to be ignored in deciding on an optimal target (in a vacuum) for the current frame. We can only choose the target we want for the current scene, and in practice if madVR smooths out the targets due to scene detection, it is what it is, it may do so in either direction. If we try to change our decision-making based on this "random" smoothing, it will just result in us overshooting or undershooting our targets (adding volatility) in other cases where scene detection does nothing or smooths in the other direction. From a statistical perspective, this is cut and dried.

Thanks, looks interesting, especially the avgHL and DPL/avgHL!
We can also take a look at the histogram.
Good idea about frame peak and frame numbers.
So this is exactly what I want(ed) to do, compile as much useful data as possible about these frames, and then additional columns for each of us and our preferred target for each of these frames, by each of us doing a manual targeting considering nothing but how the frame looks and what we prefer. It's fine or even preferable if each person has a different DPL for this, it just becomes one more useful piece of data to use for the regression, since we want to make sure it works for all DPLs. Then just try to come up with an algo to arrive as close as possible to those targets using the data only. I threw that last column in there as sort of an example. Let's say someone with DPL 800 preferred targets were close to those in that column, where they didn't want those first two frames to go above 800, liked the high targets on the 3rd and 5th, and wanted a low target on the 4th. Then the current algo works for them with a combo of DTN 75 and 2x ceiling. Or maybe their or someone else's targets are different and the formula can spit out a different combo like DTN 68 and 2.25x ceiling, or whatever. Nothing hardcoded, nothing decided, just preferred targets and a bunch of data and how can that data get us to those targets. Are you seeing the power of this type of approach after seeing those avgHL and DPL/avgHL columns?

One more thing popped into my mind: have you ever used madmeasure.exe? (It only requires the mkv as param, I'm not sure whether it works with BDMV directories.) It will create a ~100MB measurement file of the content, and madshi added tools (https://forum.doom9.org/showthread.php?t=146228) into the first post that can analyse its content (e.g. by frames), "Simple Analyzer for madMeasureHDR".
I haven't, but great idea, that seems to be possibly exactly what I want! It will probably save me from having to do so much manual work compiling the data and probably has additional data that I couldn't even get. Thanks! :)

Let's forget about the sky algo (and the rest) for now until it's fixed, because it's buggy (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60341772) (meaning unusable), (and I think it's not needed with "normal" titles, but not use).
So I just read that post you linked. I see it modified FALL, as it normally does, even on a frame where peak < DPL. But if it just reduces FALL for targeting purposes, and the resulting target is < DPL, and so then madVR just ends up with a target of DPL anyway, wouldn't that be a "no harm, no foul" situation? I couldn't see from that post if that happened or not, so just speaking hypothetically. Did it actually change the target at the end of the day?

FWIW, I did note the sky algo kicking in on some of those frames in The Grinch, which does happen to have some bright skies.

chros

12th January 2021, 17:03

Sky algo: yes, it works like that. But when it does work when it shouldn't, then what's point of ruining the content? :) That's what I meant about not usable at all. Even if it would work fine, we would need to test which value is harmless (e.g. 100 dynamic clipping is crap, and still you can set it).
So let's just disable it for now, also removing 1 variable :) So if it was on when you created the table above then it's need to be redone.

Measurement files: let's share these if one of us created them (so the others don't have to).

aron7awol

12th January 2021, 17:42

Sky algo: yes, it works like that. But when it does work when it shouldn't, then what's point of ruining the content? :) That's what I meant about not usable at all.
So what does it do in that scenario, result in a target < DPL?

chros

12th January 2021, 18:03

That's correct, resulting in target of DPL. So from the target view it is not an issue, but in reality it modifies the content, e. g. removes fine datails from it even if it's not needed.

aron7awol

12th January 2021, 18:21

That's correct, resulting in target of DPL. So from the target view it is not an issue, but in reality it modifies the content, e. g. removes fine datails from it even if it's not needed.
Are you sure it actually modifies the content? I thought it just modified the FALL used for the DTN algo calculation for targeting purposes, but didn't actually modify anything. Is the end result on-screen any different in that scenario?

Now in the dynamic clipping scenario, if it's actually modifying the content before it determines whether that's even necessary, I agree, that's a bug to me.

chros

13th January 2021, 00:00

aron7awol

13th January 2021, 00:57

Both are the same :)
Sky algo works based on histogram and yes it modifies the content (same as dynamic clipping) , that's why you have less FALL with it. How much it is visible, good question, depending on its values and content. There was a pj user who stated anything above 5 (strength) is harmful :)
Just because the FALL reported on the OSD by madVR is lower (which is the FALL it is using for target calculation) doesn't mean that it actually modified the content. In fact, can you imagine how much of an impact it would have to have on the image in order to actually reduce FALL that much! It would have to dim the sky like crazy. So I'm almost 100% sure that it just ignores the "sky" as part of calculating FALL, which is only used for target calculation. So there's no bug, it has no negative impact if it does that even on a frame with peak < DPL, because the target just ends up as DPL anyway.

So really, if I understand the way it works correctly, it can't modify the image itself, the only effect it can ever have is to potentially change the target (but never below DPL, of course), which is the point of it, anyway. It's an additional intelligence, doing something similar to what the avgHL ceiling is trying to accomplish. I think some thought it would make the ceiling redundant, but I think that both have their place based on my testing.

Which user said a strength of 5 was harmful? I'm interested in reading about their experiences. Are you sure they weren't talking about dynamic clipping?

Edit: Just tested this. It definitely only affects the reported FALL and resulting target. I was able to compare the same frame on two different profiles with the same target and one with a drastically lower reported FALL than the other, but the two images were identical.

chros

13th January 2021, 01:04

I'm sure, here (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-59183342) it is. Doesn't the histogram changes by it?
Can you ask madshi about this just in case?

aron7awol

13th January 2021, 01:10

I see no change in the histogram. Reported FALL dropped from 223 down to 151 but image is identical. I'd like to think I could notice that much of a FALL difference. :)

I don't know if that guy was confusing effects of HSTM with sky or just seeing a target difference or what, but what he reported doesn't make any sense to me based on what I've seen and what I read previously about the sky algo.

quietvoid

13th January 2021, 01:26

I'm pretty sure the sky detection is just to avoid dimming scenes that contain a "sky", where it wouldn't improve the resolved highlights in the picture by tonemapping at a higher target.
It's not supposed to change the image.

aron7awol

13th January 2021, 05:13

Yeah, and I've been running it on 100 strength for a while now and I've never seen it cause a single issue, possibly because we're pretty much immune to it causing any potential issues because our targets just can't go all that low due to high DPL.

I think it's similar to what the guys in the AVS thread are seeing with grain right now. That's only an issue at low DPLs because the curve can end up rolling through the area where that grain falls, but for us, we're pretty much always 1:1 mapping down in that area.

aron7awol

13th January 2021, 17:28

As for dynamic clipping, I agree, we don't want it working when peak < DPL, but I wonder if the logic of having it on always is that if we're toeing the line of DPL, we don't want it turning on and off within the same scene. But of course that situation could be dealt with with some additional intelligence of seeing if we're near DPL, leveraging the scene detection algo to smooth the transition, etc. My guess is madshi just didn't see all that as worth the effort and left it on always, but it would be nice for the user to be able to decide if they want it disabled when peak < DPL instead in the meantime and accept the risk of it potentially pumping on and off in that edge case.

aron7awol

13th January 2021, 20:04

So I generated a measurement file as a test to see what it contained. What do you guys use to analyze the resulting measurement files, this?

https://www.avsforum.com/threads/madvr-tool-madmeasurehdr-optimizer-measurements-dynamic-clipping-target-nits.3040072/page-17#post-57433788

That tool only seems to give limited info on frames (peak nits, LUMs, and HUEs) and scenes (scene peak), and it shows some percentiles which would be really useful if they were related to frames or scenes but instead they are percentiles of the entire file.

Do you guys know of a way to get more granular data out of the measurement file, or is this all it contains?

quietvoid

13th January 2021, 22:25

Before a tool there was an Excel spreadsheet, you can try finding that one.
Probably doesn't have new calculations or something, but you'd have access to the data.
Sadly I don't think Soulnight ever shared the code.

An alternative would be parsing the measurement files, but I don't know the format.
If you find a description of the fields, I might look into it.

I also found the other one: https://www.avsforum.com/threads/madvr-tool-madmeasurehdr-optimizer-measurements-dynamic-clipping-target-nits.3040072/page-17#post-57433788
May have more what you're looking for.

aron7awol

13th January 2021, 23:21

Thanks. Unfortunately, that's the same tool I tried :(

An Excel spreadsheet sounds good, I'll look for that.

chros

13th January 2021, 23:40

I don't know any other tool we can use for measurement files, only this one (let alone the docs for it :) ).
Basically, these were the "offline" variant of the histogram on the OSD, that's why only the brightness and hue histogram data is in it.
- on the main tab you can double click on:
-- an "avg xxx% frame peak" row (in the middle column) -> yellow graph will appear on graph
-- a scene row )on the left column to get a scene histogram + graph
--- here you can double click on a frame to get both histogram
- right click on a row -> export to txt

Can you upload the measurement file somewhere? (e.g. 7zip-ed to wetransfer.com (https://wetransfer.com/)) Thanks

Edit: I saw quietvoid already replied :)
@quietvoid: you can export it into txt (see above) if you like

Sky detection: thanks guys, then I remembered wrong, so it's not a bug :) I don't see any changes in the image here either.
(Btw, I used 700/60 because I switched off sky algo (=0).)
So, what values should we use for this during testing? 100, 20, 20, 20?

About dynamic clipping: I'm certain that is indeed a bug :) there's no hidden meaning behind it.

aron7awol

13th January 2021, 23:59

I don't know any other tool we can use for measurement files, only this one (let alone the docs for it :) ).
Basically, these were the "offline" variant of the histogram on the OSD, that's why only the brightness and hue histogram data is in it.
Hmm, so if the measurement file doesn't actually contain the frame by frame data that we would need to get the pieces of data I listed in my sheet, then I guess we're stuck compiling most of it manually (and unable to get the rest of the data that I really wanted). :( Maybe once madshi gets back on the target topic we can get him to dump some additional data on the OSD or in a measurement file.

Can you upload the measurement file somewhere? (e.g. 7zip-ed to wetransfer.com (https://wetransfer.com/)) Thanks
Sure, but I should probably do The Grinch rather than the random short episode I did just as a test :)

Sky detection: thanks guys, then I remembered wrong, so it's not a bug :) I don't see any changes in the image here either.
(Btw, I used 700/60 because I switched off sky algo (=0).)
So, what values should we use for this during testing? 100, 20, 20, 20?
Maybe the 100 10 20 20 that Neo-XP uses? I know he's done extensive testing with it, and if that's not harmful for his lower nits I'm certain it won't be for us!

About dynamic clipping: I'm certain that is indeed a bug :) there's no hidden meaning behind it.
Bug or not, the edge case I considered would be a real problem, but that's not to say we shouldn't be able to disable it when peak < DPL! After all, it's just an edge case. I'm just giving madshi the benefit of the doubt. :)

quietvoid

14th January 2021, 00:09

Hmm I would think there is frame by frame data, it was just grouped by scene to verify scene cuts.
Otherwise there's no way you can do such a graph scene by scene..
For example: https://www.avsforum.com/attachments/2019-1-18-3-jpg.2512358/

Anyways, here's the post detailing the format: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-185#post-57030812
I don't have a measurement file laying around anymore..

chros

14th January 2021, 00:34

Hmm I would think there is frame by frame data, it was just grouped by scene to verify scene cuts.
Otherwise there's no way you can do such a graph scene by scene..
For example: https://www.avsforum.com/attachments/2019-1-18-3-jpg.2512358/

Yes, and when you double click on a row on the table you will get the histogram data graphs.

aron7awol

14th January 2021, 01:28

So I decided to test Dynamic Clipping a bit, because I've never actually turned it on, and I wondered if it could be useful to us. After all, a lot of us like the sometimes pretty heavy clipping that using higher DPLs than our actual panel peak nits gives us, so I wanted to see if Dynamic Clipping 100 would be similar, stronger, or weaker than that.

I used the scene at the beginning of Valerian that I've seen others in the AVS thread use to show what Dynamic Clipping does. And I actually think, based on the testing I just did, that similar to the Sky Detection, even though the OSD always shows a change, it doesn't actually kick in when peak < DPL.

At first I was just testing at my normal high DPL, and I was thrown off because even though the frame peak was changing in the OSD, suggesting clipping was occurring, I couldn't find any change in the image whatsoever. And so after messing around with other frames and still not seeing any change, I tried a lower DPL, and sure enough, the clipping (in this case, on her necklace) was immediately in my face obvious. So then I started creeping up on DPL, and the clipping became less and less, and once I got up to the original frame peak, there was no clipping at all.

So it seems it actually works exactly as we want it to, where it only clips as much as needed to bring the frame peak down to DPL, but no more, and so if peak < DPL, not at all.

oldpainlesskodi

14th January 2021, 16:38

Sorry for being a little off topic - does anyone else's LG OLED map the color gamut to wide in PC mode? Even though the option is disabled, after going back and forwards with testing, it certainly seems that mine is mapping to wide in PC mode.

If it makes any difference, full chain in RGB, and both 60 and 23.976 fps (10bit) tested.

chros

14th January 2021, 18:03

I also use these modes on my B8, but the gamut option is (!) available in ISF Dark/Bright SDR modes, so I set it to Wide.

I used the scene at the beginning of Valerian that I've seen others in the AVS thread use to show what Dynamic Clipping does. And I actually think, based on the testing I just did, that similar to the Sky Detection, even though the OSD always shows a change, it doesn't actually kick in when peak < DPL.

At first I was just testing at my normal high DPL, and I was thrown off because even though the frame peak was changing in the OSD, suggesting clipping was occurring, I couldn't find any change in the image whatsoever. And so after messing around with other frames and still not seeing any change, I tried a lower DPL, and sure enough, the clipping (in this case, on her necklace) was immediately in my face obvious. So then I started creeping up on DPL, and the clipping became less and less, and once I got up to the original frame peak, there was no clipping at all.

So it seems it actually works exactly as we want it to, where it only clips as much as needed to bring the frame peak down to DPL, but no more, and so if peak < DPL, not at all.
Indeed!!! Thanks for this test!
I just tested the MMFR sample (keyframe at 00:19 sec, frame peak 6344 without clipping), lum max to see the blue patches.
I tried various DPLs 800/2000/4000 and set the clipping according to them: e.g. at 4000 DPL, 40 DC results in 3954 nits. If I use higher DC e.g. 100 no difference in the patches, if I go lower e.g. 39 there is difference (slight) straight away.
So this one isn't a bug either :)

Does this mean that there's a lower limit at DPL for this?
If so, then why on earth madvr allows the measured nits value to go lower than DPL?! What's the point?
I mean, with DC we clearly have the lower limit which is DPL!!! (Unlike with the sky algo where we don't have a specific lower limit, at least not that straightforward.)
What am I missing here?

(Btw, I used 700/60 because I switched off sky algo (=0).)
This one kept me thinking: are you sure that the sky algo is useful for us? (try it with The Meg (TM) sample)
Let's take these two profiles (DPL/DTN/SS):
- p1: 800/60/0
- p2: 800/85/100

The purpose of sky detection is the lower the FALL of high FALL scenes so DTN doesn't result in such high ADPLs, allowing to use higher ADPL value (85 vs 60), right?

But this one actually has a side effect:
- high FALL scenes:
-- with p1: they still have high FALL, so we don't need to use high DPL for these
-- with p2: they have way lower FALL, so to compensate we need high DPL
- lower FALL scenes (not really low ones)
-- with p1: we don't have such high ADPLs due to low DTN
-- with p2: we will have higher ADPLs (than peobably is necessary) due to the higher DPL!!!

What do you think?

Maybe the 100 10 20 20 that Neo-XP uses? I know he's done extensive testing with it, and if that's not harmful for his lower nits I'm certain it won't be for us!
I agree (if we decide it's useful for us) but that means he changed it in the last 2 months (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60250637) :)

VBB

14th January 2021, 18:39

@oldpainlesskodi - My C7 is locked to "Wide" in PC mode.

oldpainlesskodi

14th January 2021, 19:03

Thanks VBB - not just mine then. So, begs the question if auto in non-pc mode is the correct option then...

chros

14th January 2021, 19:09

Sure, but I should probably do The Grinch rather than the random short episode I did just as a test :)
Here it is (https://we.tl/t-5syLTZORR1) (I just used the biggest m2ts file), madmeasure.exe stats:

Metadata:
Mastering display luminance: 0.0001/1000, gamut: 0.68 0.32, 0.15 0.06, 0.265 0.69, 0.3127 0.329
MaxCLL: 1000, MaxFALL: 407 nits
Measurements:
Frames: 123455, MaxCLL 100%: 3916, 99.9%: 1195, MaxFALL: 884, AvgFALL: 100, AvgFMLL: 791 nits

VBB

14th January 2021, 19:18

Thanks VBB - not just mine then. So, begs the question if auto in non-pc mode is the correct option then...

Depends on what you mean by correct. Auto in non-PC mode maps to BT.709, while Wide in PC mode maps to the panel's native gamut, which is close to DCI-P3. Whatever you set the TV to, just make sure to match that in madVR. The same goes for gamma.

oldpainlesskodi

14th January 2021, 19:41

Yeah, that was kinda my point - surely in PC mode the panel should be in monitor mode and display what it thinks is being sent? After all, what is PC mode supposed to do......just curious why it maps to wide color gamut I guess.

VBB

14th January 2021, 19:50

I don't know why LG think it's a good idea to lock anything.

aron7awol

14th January 2021, 20:03

Indeed!!! Thanks for this test!
I just tested the MMFR sample (keyframe at 00:19 sec, frame peak 6344 without clipping), lum max to see the blue patches.
I tried various DPLs 800/2000/4000 and set the clipping according to them: e.g. at 4000 DPL, 40 DC results in 3954 nits. If I use higher DC e.g. 100 no difference in the patches, if I go lower e.g. 39 there is difference (slight) straight away.
So this one isn't a bug either :)

Does this mean that there's a lower limit at DPL for this?
If so, then why on earth madvr allows the measured nits value to go lower than DPL?! What's the point?
I mean, with DC we clearly have the lower limit which is DPL!!! (Unlike with the sky algo where we don't have a specific lower limit, at least not that straightforward.)
What am I missing here?
So, my quick answer off-the-cuff is that I think this ends up as just a display thing in the OSD, similar to how FALL changes from sky detection in the OSD even when nothing actually changes. But I want to think about this a little more to be sure. :)

And just for the record, I'm not sure I even want to use dynamic clipping yet at the end of the day, I just wanted to test it to see if I liked what it did, and at least it seems to work as intended and so we can evaluate potentially using it.

This one kept me thinking: are you sure that the sky algo is useful for us? (try it with The Meg (TM) sample)
I am certainly not sure! :)

Let's take these two profiles (DPL/DTN/SS):
- p1: 800/60/0
- p2: 800/85/100

The purpose of sky detection is the lower the FALL of high FALL scenes so DTN doesn't result in such high ADPLs, allowing to use higher DPL value (85 vs 60), right?
I know you mean DTN, not DPL, in that last sentence. I'm not sure I would say the purpose of sky detection is to lower FALL to allow the use of a higher DTN, as much as I'd say the purpose is really to lower FALL so that the target doesn't end up unnecessarily high in that particular scene. But that being said, that reduction of the target in those scenes may allow the use of a higher DTN in a case where a user wanted to use a higher DTN based on other scenes but only didn't because of these scenes. So I do get what you're saying, I just think that's really more of a potential side benefit than a primary intent?

But this one actually has a side effect:
- high FALL scenes:
-- with p1: they still have high FALL, so we don't need to use high DPL for these
-- with p2: they have way lower FALL, so to compensate we need high DPL
- lower FALL scenes (not really low ones)
-- with p1: we don't have such high ADPLs due to low DTN
-- with p2: we will have higher ADPLs (than peobably is necessary) due to the higher DPL!!!

What do you think?
So kind of as I refer to above, maybe we should compare the same DTN with/without sky detection first, and then evaluate whether we actually want to increase DTN on the sky-enabled profile. Because we might be happy with the lower targets it results in on certain scenes, but still like our normal targets on the other scenes. And then there's also the avgHL ceiling which can also bring down a "too high" target in other scenarios. And if there are other scenarios where we feel that a target is too high, we should try to figure out how to identify this scene from the data and maybe add some sort of additional "ceiling". So I think it comes down to whether we ever feel like we want a higher DTN based on some frames but other frames make us not want to increase it, and that's exactly where we need to focus on improving the algo.

I agree (if we decide it's useful for us) but that means he changed it in the last 2 months (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60250637) :)
I think he mentioned he changed it after switching to "don't add peak nits", for what that's worth.

oldpainlesskodi

14th January 2021, 20:29

@VBB - agreed. Oh well, more testing lol.

chros

15th January 2021, 11:49

So, my quick answer off-the-cuff is that I think this ends up as just a display thing in the OSD, similar to how FALL changes from sky detection in the OSD even when nothing actually changes. But I want to think about this a little more to be sure. :)
Ok, thanks. But you can't come up any reasonable explanation for it either then let's report it madshi, because for me it doesn't make any sense.

I know you mean DTN, not DPL, in that last sentence.
Yes, sorry, corrected, thanks.

I'm not sure I would say the purpose of sky detection is to lower FALL to allow the use of a higher DTN, as much as I'd say the purpose is really to lower FALL so that the target doesn't end up unnecessarily high in that particular scene. But that being said, that reduction of the target in those scenes may allow the use of a higher DTN in a case where a user wanted to use a higher DTN based on other scenes but only didn't because of these scenes. So I do get what you're saying, I just think that's really more of a potential side benefit than a primary intent?
Yes, that's correct/what I meant.

I think he mentioned he changed it after switching to "don't add peak nits", for what that's worth.
OK, can be, I don't remember.
So when we use it, let's use: 100/10/20/20

And just for the record, I'm not sure I even want to use dynamic clipping yet at the end of the day, I just wanted to test it to see if I liked what it did, and at least it seems to work as intended and so we can evaluate potentially using it.
...
(sky detection) I am certainly not sure! :)

So kind of as I refer to above, maybe we should compare the same DTN with/without sky detection first
Agreed: so what shall we do with these 2 during our initial testing?
I vote for:
- let's disable them (I don't think they are needed for "normal" content at all)
- we can play with them after that with the well known scenes

How does it sound?

aron7awol

15th January 2021, 17:10

Ok, thanks. But you can't come up any reasonable explanation for it either then let's report it madshi, because for me it doesn't make any sense.
I'm just trying to think about if it actually matters. What is that frame peak value used for? Smoothing algo and as a ceiling for TMing, anything else? Which value is better to use for smoothing, the original peak or the after-DC peak?

Agreed: so what shall we do with these 2 during our initial testing?
I vote for:
- let's disable them (I don't think they are needed for "normal" content at all)
- we can play with them after that with the well known scenes

How does it sound?
And this is exactly the dilemma I feel with trying to test some specific combination of settings, or with certain things disabled. They all work together to hit the targets, and so we can't really test them separately and arrive at any great conclusions, and trying to test them all together introduces so many variables.

And this is why even though it may initially sound more complicated, I actually think the simplest approach is to completely forget all the algos and all the complicated combinations of settings, and instead just look at a frame and find the target we want for that frame manually. Nothing more complicated than that. Pretend you are an AI that can analyze this frame and come up with an ideal target for that frame, what would it be? After all, if we could create the perfect algo, isn't that exactly what we would want it to arrive at?

It doesn't matter how exactly we are going to arrive at that target at the end of the day, maybe it's the existing algos, maybe it's a totally new algo, maybe it's a combination of the two. Let's cross that bridge when we get to it. We really need the targets first, to know if the existing algo can work or if it needs to be changed. Otherwise, we're just pigeon-holed into the existing algo and maybe it's impossible to make it consistently hit our targets, and so we're just stuck chasing our tail.

After all, even if you just go start testing with some combination of settings, what's going to happen when you test the first frame? You're going to look at it and decide if you're happy with it or not, right? Is this going to be a simple happy/unhappy question, or are you going to adjust up and down a bit to see if you can find something even better? In any case, inevitably you are going to come across a frame that you are unhappy with the result. What then? Are you going to then adjust DTN up/down to find something that works well for that frame? Once you do, are you then going to go back and test the previous frames that you were happy with with the new DTN and see if you're still happy with them? I hope you know I'm not being facetious, I'm genuinely asking what you are thinking as far as how to handle those scenarios during the testing.

QBhd

15th January 2021, 18:21

aron7awol

15th January 2021, 20:16

I have no idea what you two are trying to accomplish! Were have a boat load of settings and they all do different things. It's all well and good to have an understanding of what the settings do and why they are there. But there will be no perfect combination.
I feel like it should be possible to make an algo that gets very close to hitting our ideal target on a particular frame, if we just figure out what exactly drives our preference on each "type" of frame. I would expect whatever it is that makes us choose one target over another will be there in the data if we just look for it. But first we need to know what target we actually want.

Right now the targeting is really only based on 3 things:
1. FALL multiplier (but is the relationship between FALL and a preferred target even linear?)
2. avgHL multiplier ceiling (but is using >100 nits as avgHL the best value to use, and again, is the relationship linear?)
3. Sky detection to ignore flat areas so they don't disproportionately bias the FALL targeting

It works pretty well as of now, but I think it could potentially be so much better.

chros

15th January 2021, 20:29

But there will be no perfect combination.
:) And that's the issue, see above.

QBhd

15th January 2021, 20:48

So you are going to do this for ONE frame? What about the trillions of trillions of other frames?

QB

aron7awol

15th January 2021, 21:03

So you are going to do this for ONE frame? What about the trillions of trillions of other frames?
No, of course not! But we don't need trillions of frames either. We just need a small subset that hopefully covers the different types of frames (only those that are bright enough to need target > DPL) that require different targeting. This is no different from when we normally test and hone in on a particular DTN that we use other than we aren't a slave to the algo, we just need to make a note of the preferred target (nits, not DTN value) for each frame we evaluate. Really, it's just the exceptions that the current algo doesn't target so well that we really need to know about.

aron7awol

16th January 2021, 01:54

Haha @chros, I think I found a bug when both sky detection and dynamic clipping are enabled at the same time! Pretty funny. So you were kind of right about the suspected bugs, you just needed to use both together to manifest it! ;)

chros

16th January 2021, 09:35

I noticed something similar with sky strength at 199 (!) without DC, and I thought this value was 0-100 :)

quietvoid

19th January 2021, 16:09

@aron7awol I've gone looking through the avs thread again.
Here's the latest Excel version of the measurement files parser: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-215#post-57108412

Format revisions:
v4: https://www.avsforum.com/posts/57030812
v5: https://www.avsforum.com/posts/57075684
v6 (current): https://www.avsforum.com/posts/57331156/

I've started to reimplement a parser to csv just in case madshi decides it's worth it to add the black level measurements for analysis.

aron7awol

19th January 2021, 17:36

Thank you!

quietvoid

20th January 2021, 21:55

I've updated my 2018 to 05.30.10, reset all settings.
No improvement for Dolby Vision black levels, contrast 99 looks weird.

I'm questioning one thing, what do you guys have set for Color Gamut in HDR?
Auto looks less saturated than Wide, with all my inputs (FireTV stick, madVR).
But I don't know if Wide is accurate.

QBhd

21st January 2021, 01:06

I have been on Auto since day 1... but like you I am not sure if it is correct or not. I may do a few measurements this week... maybe :)

QB

quietvoid

21st January 2021, 14:05

I've read that it just increases saturation artificially, but it's weird to name an option that way.

chros

21st January 2021, 19:36

I've updated my 2018 to 05.30.10, reset all settings.
No improvement for Dolby Vision black levels, contrast 99 looks weird.
Cheers. According to Vincent, they just recently fixed the CX DV black issue with the 2nd firmware release, not before.

I'm questioning one thing, what do you guys have set for Color Gamut in HDR?
Auto looks less saturated than Wide, with all my inputs (FireTV stick, madVR).
But I don't know if Wide is accurate.
I also use Auto (Cinema preset default) with HDR10. The interesting part is that I tried it with Cinema Home preset, and "Extended" is the very same as Auto (Cinema) and not "Wide"!

I have been on Auto since day 1... but like you I am not sure if it is correct or not. I may do a few measurements this week... maybe :)
That would be nice, I only measured SDR.

I've read that it just increases saturation artificially, but it's weird to name an option that way.
Which option? :) Wide?

quietvoid

21st January 2021, 19:39

Which option? :) Wide?

Yes, Wide.

SamuriHL

22nd January 2021, 03:26

I've updated my 2018 to 05.30.10, reset all settings.

Not available for me yet. Didn't even know they put a new build out. Typical.

SamuriHL

22nd January 2021, 03:33

[05.30.10]
1. Improved disconnection in certain apps after software update

Really? Yawn.

quietvoid

22nd January 2021, 04:31

The 05.30.02 update had a lot of issues with WiFi slowing down significantly and even dropping.
They didn't mess around to publish a hotfix.
Basically these last updates just add AirPlay support.

Asmodian

22nd January 2021, 05:35

I am very happy to report that firmware 03.21.16 seems to have solved my near black banding problem when sending full range RGB to my 48" LG CX. :)

After recalibrating with the new firmware (which was needed) I am happy running 10 bit RGB full 120 Hz to the display, though I still have madVR dithering to 8 bit to reduce banding.

SamuriHL

22nd January 2021, 07:20

Where's that one finger salute emoji..... Lmao! I'm glad they fixed it for you. I just wish they'd maybe back port a few fixes. Sigh

Sent from my SM-G975U using Tapatalk

chros

23rd January 2021, 21:08

I'm glad they fixed it for you.
Better later than never :D But seriously, 9 months after releasing a product ... finally DV is maybe good now :)

You can pick up scenes as well, share the timestamps and that's it.
- let's use 800/75 then for now
-- "disabled avgHL ceiling" enabled
-- "don't add peak nits" enabled
-- all the other unnecessary options disabled (contrast/shadow recovery, dynamic clipping, sky strength)
- that means DTN will kick in at ~184 FALL (if frame peak > DPL)
- the question is what FALL/scene combo DTN is really needed and is there any scene in the title when DTN is unnecessary with these settings
-- if we can come up with a DTN value at the end for this title then that'd be the best :)

One more important thing:
- madvr can/does result in different ADPLs during playback vs still images!

The Grinch is very bright, so it's easy to find scenes like this:
- 1: 00:02:05 - 00:03:20 (2985 - 4784)
- 2: 00:06:45 - 00:07:25 (9699 - 10658)
- 3: 00:13:10 - 00:14:00 (18930 - 20130)
-- this scene is interesting, there are frames when frame peak is only ~430 but FALL is >200.
- 4: 00:16:00 - 00:16:45 (23006 - 24086)
So, I had a little time to play with these (I added frame numbers to scenes as well), in short:
- I disabled (!) DTN for "normal" content (in my default HDR10 profile)

Here is why, DTN cause more harm than good with "normal" content:
- I haven't seen any frames that needs DTN (dimming)
- madvr works with scenes:
-- it resets on scene changes
-- within a scene it slowly corrects itself
--- often happens that there are high FALL frames for only couple of seconds that still unnecessarly dim the following frames in the same scenes

The real question is what contents are "normal" for us high DPL users, based on avgHL or frame peak or something else (as we already talked about it). I bet 95% of the content we come across today are like this.

What clear now is that the current implementation of DTN is completely wrong for high DPLs, meaning we definitely can't use 1 profile settings for all the content we have.

SamuriHL

23rd January 2021, 21:38

LG hasn't given me a whole lot of reasons to upgrade my C8. Lower max nit level. Still screwed up DV (may be fixed finally?! WOO?). PC mode still not 100% awesome. I mean they're getting better but I'd need a flawless panel to get me to want to upgrade. I want to see what the hell they've done this year because the Cx (not to be confused with the CX LOL) line didn't get the upgraded panel tech this year. WTF? They really know how to confuse people.

chros

23rd January 2021, 22:41

I agree, and not to mention new panel issues: horizontal banding, grid pattern, etc.
For me the biggest issue with separating Gx from Cx is that these will have different firmwares! What a mess this will be :D

Asmodian

24th January 2021, 00:36

I am hoping all models get the new panels in 2022 and re-unify the Cx and Gx.

PC mode still not 100% awesome.

I agree that it isn't 100% awesome but the newest firmware on the 2020 models solved my last significant issue with PC mode. Banding seems similar in PC mode v.s. HDMI mode now, in both HDR and SDR.

I can always get more picky, and there are plenty of imperfections when I do, but PC mode has improved so much since my C7 that I have to be happy about it. At least for a bit. ;)

aron7awol

24th January 2021, 00:54

So, I had a little time to play with these (I added frame numbers to scenes as well), in short:
- I disabled (!) DTN for "normal" content (in my default HDR10 profile)

Here is why, DTN cause more harm than good with "normal" content:
- I haven't seen any frames that needs DTN (dimming)
- madvr works with scenes:
-- it resets on scene changes
-- within a scene it slowly corrects itself
--- often happens that there are high FALL frames for only couple of seconds that still unnecessarly dim the following frames in the same scenes

The real question is what contents are "normal" for us high DPL users, based on avgHL or frame peak or something else (as we already talked about it). I bet 95% of the content we come across today are like this.

What clear now is that the current implementation of DTN is completely wrong for high DPLs, meaning we definitely can't use 1 profile settings for all the content we have.
Did you have sky detection turned on? I found DTN to not do very much with The Grinch with it on, which sounds like what you want it to do.

chros

24th January 2021, 01:17

It was off as per instructions along with the others. My point is that based on those 4 scenes TG doesn't need DTN!

aron7awol

24th January 2021, 01:42

It was off as per instructions along with the others. My point is that based on those 4 scenes TG doesn't need DTN!
Okay, but sky detection seems to work perfectly on that movie to lower FALL so that DTN does little/nothing, which seems to be exactly what you want on those scenes. Otherwise, you have to lower DTN in order to stop it from kicking in and see how that lower DTN works with other content.

Sky detection is designed to work hand-in-hand with DTN (it modifies FALL only for the DTN calc and does nothing else), so I think it's premature to conclude that the current implementation of DTN is completely wrong for high DPLs when we've only tested a single DTN and only without sky detection doing its job to lower targets in flat skies which are abundant in The Grinch.

I think if you enable sky detection on those scenes you will come to quite the opposite conclusion, that the targeting actually ends up very close to where you want it on those scenes with the 75 DTN.

chros

24th January 2021, 12:32

Okay, but sky detection seems to work perfectly on that movie to lower FALL so that DTN does little/nothing, which seems to be exactly what you want on those scenes. Otherwise, you have to lower DTN in order to stop it from kicking in and see how that lower DTN works with other content.
Yes, it helps, bit still not perfect, DTN is triggered once in scene 2 (although just for a short time), but in scene 4 similar thing happens than without it:
- "disable avgHL ceiling" disabled
- avgHL ceiling: 2x
- sky strength: 100

that the targeting actually ends up very close to where you want it
And that is my point: it doesn't. I think TG (that is an extreme example of "normal" contents) doesn't need DTN at all.

So, if someone comes across these type of contents (e.g. maxCLL < 1300-1600 nits) that benefit from DTN then please share it (with timestamps).

aron7awol

24th January 2021, 13:53

So let's focus on those scenes where it is doing what you don't want it to do, even with sky detection enabled.

For each of those scenes make a note of:
With sky detection enabled, what is FALL, and what is the resulting target?
If you lower avgHL multiplier further, is there a point where it kicks in?
Does changing the sky max width 10 vs 20 or the hill settings help ?

aron7awol

24th January 2021, 22:25

https://i.imgur.com/KA8mV6X.png

Did some testing of those same scenes to see what impact the avgHL ceiling and sky detection would have on those frames which result in targets >800 with just DTN 75:

1. Sky detection lowered the target down to 800 in 4 out of the 6 frames.
2. In the two remaining frames, a 2.5x avgHL ceiling lowered the target down to 800.
3. Thus, running sky detection and a 2.5x avgHL ceiling resulted in a target of 800 in all of the frames tested.

I went back at the end to one of those two frames (6:59 shown at the bottom of the list) where sky detection didn't lower the target, and tested Dynamic Clipping to see if it would have any impact. A value as low as 20 lowered the target down to 800 alone.

So I think this is a good exercise showing the power of the ancillary functions of the algo, where we see each of them able to chip in and lower targets in certain cases, and although one of them alone didn't do that completely, when enabled together they can combine to give us the targets we want. And let's keep in mind we haven't even established that 75 DTN is what we want, but it still shows that DTN can work for this particular movie with sky and avgHL ceiling enabled.

I added a column to the far right showing the maximum DTN for each frame that would result in a target of 800. I think that columns shows nicely that relying on just a single DTN FALL multiplier to end up at the right targets is an exercise in futility, and that some ancillary functions are essentially required in order to get good targets. I don't see this as a knock on DTN, it's just that FALL alone only tells us so much, and so the FALL multiplier is just one piece of the puzzle.

So @chros, I think if you run through The Grinch with sky detection on and a 2.5x avgHL ceiling (I'm leaving DC off for now because of the order-of-operations error when used concurrently with sky detection) you'll be very happy with the resulting targets. The question then becomes, are there any frames where that combo results in a target we aren't happy with? We now need to find those so we can take the next steps.

Edit: I went back through with those settings and found a frame at 2:20 which resulted in a target >800. The target was 830, which was avgHLx2.5 and so avgHL was 332 for this frame. Lowering DTN to 65 lowered the target to 800. Additionally, I switched back and forth between the 800 and 830 target for this frame and really couldn't tell any difference anyway.

Edit2: Also forgot to mention that 2.75x works almost as well as 2.50x to keep those targets down. It resulted in some that were something like 806 nits instead of 800 but that's inconsequential. Just wanted to mention it because 2.50x might be a little lower than we want for some other stuff like The Meg. FWIW, my previous testing had found an avgHL multiplier that was approximately 70% of FALL multiplier seemed to work well, which in this case would be 3.05x for the 800/75 combo. That is the exact combo I liked a lot with The Meg: 800/75 3.00x. This is good, I think, that combo of settings actually works quite well on both The Grinch and The Meg. I might choose something in between 2.75x and 3.00x if that were possible.

chros

25th January 2021, 23:20

Thanks for the detailed post, but I got different results than you, how is this possible? :)
Here are the 2 frames (that cause the issues in scene 2 and 4, you can you ctrl+g in mpcbe to jump there (note that directshow isn't not frame accurate seeking), and btw there's a new repeat A-B feature in the latest mpc-be beta (under Play -> Repeat) that is really useful for testing scenes :) )

Here's my settings (https://imgbox.com/1bCmZUqW) (note that I still use b119, not the newest):
- note I used my SDR monitor for reading data from it (not image comparison, for that I used the TV)
-- calibration is set to bt2020 (although just the measured peak changes with e.g. bt709)
- lum method and desat options doesn't seem to change anything
- max width of sky strength doesn't seem to change anything either (using these 2 frames)

1. @7:12 frame 10361: frame peak 990 (vs 977), FALL 321.4 nits (vs 324)
- 2.5x avgHL: 990 (vs 977)
- sky100: 823 (vs 800) (FALL 188.62)
- 2.5x avgHL + sky100: 823 (vs 800) (so sky detection does the job here alone)

Now I wonder where the different values come from in your vs my case...

2. @16:17 frame 23434: frame peak 809, FALL 465 nits
- 2.5x avgHL: 809
- sky100: 809 (FALL 443.49)
- 2.5x avgHL + sky100: 809

This is the first frame of the scene, it's seemingly harmless, but that's (and the following high FALL frames) the cause for dimming other frames unnecessarely here.

aron7awol

26th January 2021, 00:03

What is the XXX (vs YYY) referring to?

I think the difference in our results are due to two things:
1. Different frames (I didn't have exact frame numbers)
2. Scene detection (I always reset it for these purposes)

Looking at your results on those two specific frames...
1. So sky reduces FALL to 188.62 and results in a target of 823. Is the difference between 823 and 800 noticeable here? If so, reducing DTN to 73 would bring the target down to 800 here.
2. This frame looks fine to me. 809 vs 800? I don't think we need to be worrying about that, it won't be noticeable.

If future frames get higher targets due to previous frames, that's a scene detection issue, not a targeting issue. We can't worry about that when working on targeting, if we want to fix that we need to work on scene detection.

Hopefully the new FALL features in this new build work great!

Edit: FALL detail threshold of 4 lowers almost all targets to 800!

quietvoid

26th January 2021, 05:20

This 122 build is definitely going to throw you guys off :D

I've gone from DTN 110 to 200 + detail threshold 3.
My new settings: https://ibb.co/yyzzBy6

Definitely improves brightness in The Meg, but it's a bit worse in terms of highlight contrast in S&M.

chros

26th January 2021, 11:18

What is the XXX (vs YYY) referring to?
XXX: values here, YYY: your values from the table. These should be the same.

I think the difference in our results are due to two things:
1. Different frames (I didn't have exact frame numbers)
2. Scene detection (I always reset it for these purposes)
I don't think so:
1. I stepped back/forth (with ctrl+ left/right arrow) and I couldn't find your frame
2. I also reset scene and not just stepping forward

I think it's the source we use :) I use the untouched full bdmv.
If you use a reencoded one, that could explain the difference in FALL (although I don't see why the frame peak would be different).
If you like, I can cut out clips and upload it to be able to use the same thing. Otherwise we just talk about apples vs oranges :)

Looking at your results on those two specific frames...
1. So sky reduces FALL to 188.62 and results in a target of 823. Is the difference between 823 and 800 noticeable here? If so, reducing DTN to 73 would bring the target down to 800 here.
2. This frame looks fine to me. 809 vs 800? I don't think we need to be worrying about that, it won't be noticeable.
I'm not sure about this, but let's find out first why we have different values.

If future frames get higher targets due to previous frames, that's a scene detection issue, not a targeting issue. We can't worry about that when working on targeting, if we want to fix that we need to work on scene detection.
Hmm, in clip 4 that scene is a big panning shot that's why scene detection doesn't catch it.

Hopefully the new FALL features in this new build work great!

This 122 build is definitely going to throw you guys off :D
:D More options :) But I don't want to deal with them now, it's complicated enough already :)

SamuriHL

31st January 2021, 03:21

I agree with the complicated! I've not had a chance to even touch the latest build yet. Currently my entertainment center is undergoing a massive transformation. I've finally decided to wall mount my 2 65" tv's. The C8 in the living room on a fully articulating arm and the beast known as the tc-p65vt50 panny plasma is being wall mounted in the bedroom. That's the one that started this project because I noticed the stand I have it sitting on (5' above the floor) is bowing in the middle from the weight of that thing. UGH. It is also going on a mount with an articulating arm that can handle the weight. Not as nice as the living room mount but still awesome for a bedroom tv. Both entertainment centers are getting replaced with audio racks at some point. Eventually the C8 will replace the panny plasma in the bedroom. Just as soon as LG makes something worthy of replacing the C8 with. :D

In any case, once I get everything back together I'll be testing again. I have a week of vacation coming up soon for my b-day and will have lots of time for movie watching. I'm hoping you guys can figure out some starting point settings to work on by then. LOL

quietvoid

4th February 2021, 20:42

Might try the new curves stuff this weekend, but it's pretty intimidating with the little info madshi gave so far.
Maybe I'll wait for others to dip their toes in first :D

At least if there's a working build too.

SamuriHL

4th February 2021, 23:14

I was thinking the same thing. Holy hell. Power? Yes. But that's going to be quite a bit of work. I have no idea where to begin here.

quietvoid

5th February 2021, 06:08

Welp, it looks like we're back to the old non-DTN clipping for HDR output :(
So until that's fixed, I don't think we can test curves correctly..

chros

5th February 2021, 19:06

Since we started again to talk about DTN I watch approx 30 titles (TV shows/movies) and I just recently came across 2 recent titles that might (?) be good candidates for DTN testing:

Elysium (2013):

Metadata:
Mastering display luminance: 0.0001/1000, gamut: 0.68 0.32, 0.15 0.06, 0.265 0.69, 0.3127 0.329
MaxCLL: 2594, MaxFALL: 1698 nits
Measurements:
Frames: 157656, MaxCLL 100%: 2449, 99.9%: 2093, MaxFALL: 1688, AvgFALL: 42, AvgFMLL: 631 nits

This one has pretty high nits, but the avgFall is average at best, I don't know where that insane 1688 MaxFALL can be. FALL sometimes goes up to 140 nits.
The problem is that I don't know where/when these frames are.

The other one is 2012 (2009):

Metadata:
Mastering display luminance: 0.005/4000, gamut: 0.68 0.32, 0.15 0.06, 0.265 0.69, 0.3127 0.329
MaxCLL: 4451, MaxFALL: 514 nits
Measurements:
Frames: 227064, MaxCLL 100%: 5145, 99.9%: 2352, MaxFALL: 513, AvgFALL: 9, AvgFMLL: 266 nits

Here there are only 4 instances where frame peak goes above 2000 nits, and the avgFALL is only 9 nits :D So I guess this is not a good example after all :)

Here are the 2 measurement files (https://we.tl/t-i4mVGemFSU) in one.

@aron, did you managed to modify the excel calculation to have FALL values per frame? (That's what I miss from the madMeasurementAnalyzer tool.)

Might try the new curves stuff this weekend, but it's pretty intimidating with the little info madshi gave so far.

Yes, seems pretty complicated.
I'm not sure where those 110, 509, 2230 and 10000 frame peak nits values come from.
But in our case, with >=700 DPL, I guess the first 2 must set to clipping, and only the other 2 needs to be modified.

Welp, it looks like we're back to the old non-DTN clipping for HDR output :(
Thanks for noticing this, the old bug is back :)

aron7awol

5th February 2021, 20:36

Yes, seems pretty complicated.
I'm not sure where those 110, 509, 2230 and 10000 frame peak nits values come from.
I think he first chose 110 as a reasonable value to draw a line at. Then he just divided the remaining EOTF by 3:

51.78% -> 110 nits
Split the rest of the way into 3 and we get...
67.85% -> 509 nits
83.93% -> 2230 nits
100.00% -> 10000 nits

chros

5th February 2021, 20:50

Hmm... You are really good in this! :)

aron7awol

6th February 2021, 00:08

As for the other testing stuff, this custom curve stuff just opened a ginormous can of worms! Everything was already as clear as mud. Now I'm just gonna sit back and see how this custom curve stuff shakes out before trying to do anything else. :)

SamuriHL

6th February 2021, 01:37

Yea I don't even know where to begin with all the options now. Like at all.

quietvoid

6th February 2021, 02:49

I might look at it at high peak in SDR, but I'm not sure it's going to be useful since 1:1 mapping for SDR would probably look bad.
So waiting for a non-DTN fix. I don't think the curves worked at all with DTN on and HDR, but maybe I have to lower it a lot.

Edit: I don't really know what I'm doing but it's definitely worse than default so far.

quietvoid

6th February 2021, 17:43

Turns out it's fine to use DTN with the custom curves, the difference is just much less noticeable.
I made a curve that helps with some highlights (Mad Max, BvS) but it also clips some stuff in the S&M bench.
Still have to figure out how to fix the clipping, but otherwise it's not that bad.

If anyone wants to try: https://forum.doom9.org/showpost.php?p=1935512&postcount=1516
I used 870 DPL to test, in SDR seems to be similar behavior so that's easier to test without a TV.

quietvoid

8th February 2021, 04:23

To stop updating the mega link I've made a gist for my curve settings.
https://gist.github.com/quietvoid/cf1454ad05efba1ad34cf6963b55ca1a

madVR 123b settings
https://i.postimg.cc/JHC9yJZx/madvr123b.png (https://postimg.cc/JHC9yJZx)

For some reason using color tweaks affects brightness a lot so I disabled it.
Since there's some brightness decrease with the curve in places, lowered the desat a little to compensate.

So far I can't find anything very wrong with it.
I still have to play with high brightness demos for the x8 and above curves, I think.

SamuriHL

8th February 2021, 15:53

Kick ass. Thanks for that. The audio racks are scheduled for delivery today and I'm on vacation next week so I'll have more time to play next week. I can't wait to try your settings though since I'm using 870, as well, and we have the same TV I think they'll work well for me.

Magik Mark

9th February 2021, 05:02

Thank you for sharing QuietVoid!

quietvoid

15th February 2021, 22:30

Had a quick look at 124, I decided to reenable contrast recovery because the curves flattened things quite a bit.
Protect top seems to help a little but otherwise I haven't seen a difference with protect bottom.

Not sure what to think about the alternative HSTM, some of you guys might like how it increases overall image brightness but there's also a lot of contrast loss, in S&M at least.
Like it totally changes the contrast, but also helps keep brightness in some of The Meg skies.

Also got rid of the desat 2.

124 settings now: https://i.postimg.cc/NFFVJycd/madvr124.png
Curve is still whatever is on the gist, haven't updated.

SamuriHL

16th February 2021, 01:32

Did you try Fer's medium contrast recovery? I haven't had time yet to play with these things. Hoping tomorrow I'll have a few minutes.

quietvoid

16th February 2021, 02:54

I haven't tried higher strengths, I'll have a look.
From a quick glance it seems better at keeping contrast and brightness, so it might be better.

SamuriHL

16th February 2021, 03:48

I'm hoping I'll get to some testing tomorrow. I have the week off but I do have a few things to get done earlier in the week. I'm excited to play around with these curves. Totally sure I have no real idea how they work but that's never stopped me before. LOL

SamuriHL

16th February 2021, 21:55

I'm using your settings with Fer's old medium. I like what I'm seeing. A lot.

quietvoid

16th February 2021, 21:56

It's too bad I don't understand why shadow detail isn't being recovered by the curves at high nits, apparently everyone else has improved shadows.

SamuriHL

16th February 2021, 23:48

That's a good question. That would be a nice improvement. I haven't actually looked at how you defined the curves yet. I just pulled them in and watched a movie. It looked damn good.

quietvoid

17th February 2021, 05:17

I'm getting quite tired of testing my own curve, and it doesn't seem to be optimal.

I realized testing in SDR didn't really show properly how the brightness would differ, and I was trying to retain more brightness with the curves.
Turns out everyone is just making the curve as flat as possible to retain detail, and using other settings to get some brightness back.

In HDR, the brightness difference really is not very noticeable but the detail retained is.
So I'll just settle with Neo-XP's curve adapted to 870 DPL, until madshi comes up with something more automated.

I posted about issues with Fer15's medium contrast recovery, so I'm sticking with Neo-XP's Log Conservative even though it is slightly less bright.

To get back the shadow detail it's necessary to use the alternative histogram, and though it increases overall brightness there's not that much contrast loss in HDR.
It's hard to perceive if an overall darker image with the same highlights looks more "contrasty" compared to a flatter background, but that's the compromise with all curves so far.

So for now I updated my gist to Neo-XP's adapted curve, and these are my settings: https://postimg.cc/PLPwdwG2

chros

17th February 2021, 11:22

@quietvoid, can we playback single layer (profile 5/8) mkv (!) DV files with FireStick 4k? Of so, how? Thanks
(I tried to remux it into ts with latest tsmuxer nightly, but oppo doesn't trigger DV with those files. And I don't know how to extract hevc+rpu to create an mp4 DV file with dbl mp4muxer.)

quietvoid

17th February 2021, 14:49

Oh man you made me post out of topic instead of linking to makemkv :(

Anyways now I'm curious if my 200 DTN can be lowered.
With Neo-XP's curve, disabling DTN results in way less clipping than the default curve.
Now DTN can actually just be used for contrast :D

I ended up lowering DTN to 100, so back to what it was before.
Except with the curves now there is no clipping at all.

madVR has never looked this bright AND detailed, this is what HDR is supposed to look like.
Not the dark, high DTN look. At least IMO.
We've just been stuck with going darker because the curve wasn't optimal to retain detail.

Most shots in S&M keep their contrast, but everything else is just way brighter and midtones/shadows are improved a lot.

SamuriHL

17th February 2021, 17:13

Damn. I'm going to give this a try later today. Post a new screen cap of your madvr settings?

quietvoid

17th February 2021, 17:19

Not much changed except for DTN: https://postimg.cc/7G7FDhGt
Using Neo's curves adapted to 870, the file is in my signature.

If you like things darker either increase DTN or try disabling the alternative histogram and/or FALL detail (those are good to keep bright skies).

SamuriHL

17th February 2021, 18:18

In my environment I like things slightly brighter. I use a lower DTN than you do (or did before this latest round of changes). I'm going to play around this afternoon. This should be fun.

SamuriHL

17th February 2021, 19:00

Well this is definitely getting us somewhere for sure. I think we need to tweak a bit here and there but this is the right direction to be going in!! I like!

quietvoid

17th February 2021, 19:54

We might have to look again at how the TV behaves above 870 nits target again, too.

SamuriHL

17th February 2021, 19:57

That's a good point. A lot of testing in our future.

chros

17th February 2021, 21:50

Thanks for the reply, quietvoid!

Back on topic: although I haven't played with the curves yet (DTN is off here), but how does it affect (along with the rest of the options) the resulting PQ curve?
I mean, if the image is brighter then what happens to director's intent? :)

quietvoid

17th February 2021, 22:13

The curves by themselves don't make anything brighter, they just resolve the highlight detail better.
I compared passthrough and DTN 100 and they seem to have same brightness midtones, so I'd assume it's fine.

Scenes under 509 nits are set at 1:1 and should be left untouched.

Now if you don't use DTN I don't think everything becomes brighter, but you lose contrast in those high brightness areas.

SamuriHL

18th February 2021, 00:40

I went a little too hot when I watched Infinity War earlier and had DTN set to 75. I need to back that down a bit. I think 100 maybe a good place to be. One of the things that impressed me is the tesseract in the beginning. It's one of the reasons I use Infinity War to judge how these things help or hurt. They aren't especially bright scenes, but, there are points in there that are very bright, plus fire, plus a ton of shadows. But the tesseract is something I pointed out to madshi before. It can NOT be looked at statically. I.E. freezing it won't help. But when you have a proper curve and settings, the DETAIL moving within that thing as Loki holds it up in front of Thanos is unbelievable. There's shadow detail within the damn thing and it moves. Your settings with Neo's modified curves are some of the best representation of that I've seen in a long time. I like it a lot.

chros

18th February 2021, 11:41

I compared passthrough and DTN 100 and they seem to have same brightness midtones, so I'd assume it's fine.
Which content did you test it with? Maybe I'll play around with these during the weekend.

quietvoid

18th February 2021, 14:17

I mostly test with S&M, Batman v Superman and The Meg.
In this case I think The Meg has similar levels with alternative histogram and FALL detail brightening skies quite a bit.

I think it's possible that madVR is trying to make things as bright as the source PQ values would allow, hence matching brightness with passthrough.
Obviously with the exception of highlights, where the tonemapping is much better (and the only place where creator's intent might be different, because of garbage default TV mapping).

By the way, I quickly looked at 1000 target nits and it still seems to be clipping, so nothing changed for that I guess.
Maybe the curves had to be readjusted but I doubt it would do anything.

chros

18th February 2021, 19:32

Thanks for keeping the spirit alive :)
I just tried out your settings and curves (@800 DPL) with MMFR sample: it doesn't fix the big blue patches in the beginning.

Also, does the DTN work at all with b124? (the bug you have reported with b123) I don't think so ...

quietvoid

18th February 2021, 19:44

The curves are not meant to fix the "issues" in MMFR, that's a saturation problem.

DTN works just fine, with the default curves it would just clip, but now there is more detail without DTN as well.

If you didn't regenerate the curves for your DPL there might be differences, the file only has points for 870.
You'd have to follow Neo-XP's method to redo it for another DPL.

SamuriHL

18th February 2021, 21:03

I've dialed DTN to 85 for my particular setup. It seems to have a good balance for me. Quite impressive really.

chros

18th February 2021, 23:00

Are you sure that DTN works? ADPL value does not change (it's 800 all the time in brackers) if curves are disabled.

SamuriHL

18th February 2021, 23:48

I don't know if it "works" but it certainly has an impact on brightness.

quietvoid

19th February 2021, 02:46

Are you sure that DTN works? ADPL value does not change (it's 800 all the time in brackers) if curves are disabled.

It definitely works on my machine.

Without curves: https://slow.pics/c/RCrIeQ2A
With curves: https://slow.pics/c/9C0PpnI2

chros

19th February 2021, 08:27

You also only got 870 for ADPL. Dunno about with curves, but without curves it should change (increase).

quietvoid

19th February 2021, 14:46

What even is ADPL? The value in the brackets is increasing with DTN enabled..
You can click to change image.

chros

19th February 2021, 21:02

Oh, I missed that, sorry. It works here as well, thanks, I should have checked with SM straight away.

So, what does tge new clipping at mastering peak means to us (that is enabled by default) ?
There are lot of contents which have master 1000 nits but maxCLL is above 2000 nits.
Does it mean that madvr now clips at 1000? If so then there's no need for any algo for these contents for sure,

quietvoid

19th February 2021, 22:39

You can always compare by reencoding with different metadata.
I assume it's to respect the grading intent.

SamuriHL

20th February 2021, 00:53

If only there was a way to override the metadata LOL In my new setup I don't have the Integral 2 connected right now. I don't have a good way to connect it permanently to the system but that is a very good device to have for this kind of stuff.

chros

20th February 2021, 11:51

Nah, too much hassle, I'll ask him about this instead. :)

Back to your settings/curves, couple of observations (@ 870 DPL, I used this to have the same setup, although the diff should not be huge at such high DPLs):
- out of these samples (https://mega.nz/folder/n5dzgQbJ#A99MVeafyffB1qK6-ahrTg) only The Meg (03*tm.mkv) increased the DTN during normal playback
- take a look at the 10*pr.mkv @00:00:22 sec (keyframe):
-- your settings (I think) unnecessarily dims the image, reducing highlights
-- metadata: MDL 4000, maxCLL/FALL 2465/823
-- frame stats: peak 1710, Fall 32.840 (with your setting 19.401) resulting ADPL 870

So, as far as I understand how this works (please correct me if I'm wrong):
- DTN with curves only selects (interpolates between) the right curve(s):
-- based on frame peak (what about the new mastering peak clipping feature?) and resulting FALL
-- if DTN is disabled with curves, only the x1 curves are used
- with 800/870 DPL, only the x1/x2 curves will be used with 2230 (neither the 4x/8x/16x ones), because x2 means 1740 ADPL with 2230 which is really high already
- the frame above uses 2 curves: 509x1 and 2230x1 (but closer to 2230x1)
- the only diff with 2230x1 between your curve vs default is the custom knee: 0.50 vs 256! (that makes the curve less steep, reducing brightness)

Why do we need so small knee with our high DPL? Neo used the 0.50 knee with 100 DPL :)
So I revert back the 2230x1 curve to default and maybe the 2230x2 as well (although I haven't thought about it yet).

quietvoid

20th February 2021, 17:20

I wouldn't say that frame at 22s is unnecessarily dimmer, some of these lights are noticeably closer to being blown out with the default curves.
You can just make it brighter with dynamic clipping at 50 (for this specific scene), without affecting the detail.

This is also so close that I'm not sure you'd appreciate the difference in HDR playback.

I don't believe (anymore) that curves should be used to increase brightness, because you're definitely going to lose detail that way.

SamuriHL

20th February 2021, 17:21

I tend to agree. Use the curve to retain as much detail as possible. Brightness vs detail is the compromise, as madshi has said numerous times. You give up one to have the other as they are on opposite sides of the spectrum. TV algorithms it's pretty clear to see the choice they make, unfortunately.

chros

20th February 2021, 20:03

You can just make it brighter with dynamic clipping at 50 (for this specific scene), without affecting the detail.
This was a really good idea, thanks, I updated the post over avsforum:
"dynamic clipping has still an effect on the resulting curves/image (tested with MDL 1000, frame peak before dc 1621 and after dc 1186, so it's above MDL), so it clearly doesn't clip at MDL."

I wouldn't say that frame at 22s is unnecessarily dimmer, some of these lights are noticeably closer to being blown out with the default curves.
"Closer" means they are still not, so why degrade the image? :)

This is also so close that I'm not sure you'd appreciate the difference in HDR playback.
That's true but this can be said about all the other options as well, even with sources that we know, let alone those that we watch for the first time :)

I don't believe (anymore) that curves should be used to increase brightness, because you're definitely going to lose detail that way.
At some point, yes, as we can see with the meg and SM, but the question remains: which is that point? :)

The curves are not meant to fix the "issues" in MMFR, that's a saturation problem.

Err, I think these curves are supposed to fix the saturation issues at the first place, although MMFR is an extreme example.

SamuriHL

20th February 2021, 21:32

No, madshi said saturation will be revisited after the curves are selected. There are people who are choosing to have desaturation options still enabled for some content so the curves are not going to fully fix saturation. They will have an impact, certainly.

aron7awol

21st February 2021, 06:06

So, as far as I understand how this works (please correct me if I'm wrong):
- DTN with curves only selects (interpolates between) the right curve(s):
Agreed, and I guess I'd say this is sort of all DTN ever did, was set the BT.2390 curve used based on peak and the resulting compression of target/DPL.
-- if DTN is disabled with curves, only the x1 curves are used
Agreed, as without DTN there is no compression.
- with 800/870 DPL, only the x1/x2 curves will be used with 2230 (neither the 4x/8x/16x ones), because x2 means 1740 ADPL with 2230 which is really high already
Well, a frame with peak and target >1740 would result in compression >2x so it should interpolate using the 2230x4 curve as well in those scenarios. And then even as we go further to frames and targets >3480 it would use that 2230x4 to interpolate along with 2230x8 (and 10000x4 and 10000x8), and even more extreme cases with frames > 6,960 we would interpolate between 2230x8, 2230x16, 10000x8, 10000x16. So really, almost all curves can be used with almost all DPLs in the right scenarios.
- the frame above uses 2 curves: 509x1 and 2230x1 (but closer to 2230x1)
Agreed, since DTN doesn't kick in, there is no compression and so it only needs to interpolate between 2 curves. As soon as DTN kicks in (and thus results in compression) we interpolate between 4 curves, except in the coincidental cases where peak and/or compression happen to land exactly on one of the options. :)

chros

21st February 2021, 13:04

Well, a frame with peak and target >1740 would result in compression >2x so it should interpolate using the 2230x4 curve as well in those scenarios. And then even as we go further to frames and targets >3480 it would use that 2230x4 to interpolate along with 2230x8 (and 10000x4 and 10000x8), and even more extreme cases with frames > 6,960 we would interpolate between 2230x8, 2230x16, 10000x8, 10000x16. So really, almost all curves can be used with almost all DPLs in the right scenarios.
Yes, you are right, sorry, I meant titles under that 2230 curve.

Agreed, since DTN doesn't kick in, there is no compression and so it only needs to interpolate between 2 curves. As soon as DTN kicks in (and thus results in compression) we interpolate between 4 curves, except in the coincidental cases where peak and/or compression happen to land exactly on one of the options. :)
About this: I wanted to test elevated black with Mehanik's (https://drive.google.com/uc?id=1bAtIc0C2GCZZVbgV-ntSKED4q-9C8L5_&export=download) black clipping pattern (01. Black level\01. black-level-v1.mp4), but the image doesn't change at all when I modify the 110x1 curve!!!
Can you guys reproduce it? (either in SDR or HDR ouptut)
Video stat: frame peak 11 nits, FALL 0.051.
Curve 110x1 mod: I moved up the first 11 (!) points to around 10 nits!
There's no change in the image at all when you switch off/on the curves.
What's happening here?

Edit: it doesn't change either if you set all the points to the max (e.g. by clicking the "create bt.2390 curve now" with its default settings).

Lastly, as he stated, madvr now should clip at MDL straight away (although it doesn't seem to work now).
But if that will be the default behavior then there won't be need for DTN at all for MDL 1000 titles!!! (which I guess the 70% percent of all the content today, at least)

quietvoid

21st February 2021, 17:29

Lastly, as he stated, madvr now should clip at MDL straight away (although it doesn't seem to work now).

I had posted an example where the MDL clipping worked: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-568#post-60401221

quietvoid

21st February 2021, 20:13

Anyways I ended up making a minimum steepness + half max knee file, seems to be better than what I had done for Neo-XP's adapted one.
It's in my signature.

Alternatively, here's Neo-XP's 1.2 curves adjusted for 870 DPL: https://gist.github.com/quietvoid/d797c47827d39651822d21aa56b77beb
It's a bit more contrasty/darker, see https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-607#post-60518938

Since it lowers brightness a bit more, I lowered DTN and increased the FALL detail to gain back some brightness. But again that's totally subjective.
Settings: https://postimg.cc/n9Fr9s7r

Now I need to start actually watching stuff.

chros

21st February 2021, 20:28

I had posted an example where the MDL clipping worked: https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-568#post-60401221
Omg, that was a month ago and I almost forgot about it :) Thanks
Yes, that option works in b123 but not with b124/125.

Anyway, can you guys experiment with that option at the MDL value if you will have time whether it's needed or not? I'll do the same.
Because if he fixes it there won't be turning back ... :)

Anyways I ended up making a minimum steepness file, seems to be better than what I had done for Neo-XP's adapted one.
It's in my signature.
Cheers, I'll take a look some other day.

Now I need to start actually watching stuff.
:D Sure

PS: I gathered up the madvr beta changlog (https://pastebin.com/6N9tr0t6) into 1 file from b100 (I haven't gone back more, yet?), I'll plan to update it in the future (I've also put the link in the post of the link in my signature).

jigar24

26th February 2021, 05:58

My 1080p bluray player (couple years old) has options for 4:2:2 ycbcr, 4:4:4 ycbcr, or RGB. With the UHD players I believe they offer those choices too, along with up to 12bit output. By default on 'auto' pretty sure they all output ycbcr at either 4:2:2 or 4:4:4

quietvoid

7th March 2021, 07:16

chros

7th March 2021, 11:24

Not too sure how much the new normalization affects brightness, I haven't compared previous versions.
This should be an option, again, you can't watch content normally to actually see what the curves do, unless you revert back to b128 (but there's the mdl clipping :) ). That's why I asked him to clarify that.

So, for all of these that happened in the latest builds, I went back to b113, until everything settles down :)

Different question: did any of you calibrated DoVi? If so, do you still have the DoVi config file that can be shared?

chros

7th March 2021, 11:52

Different question: did any of you calibrated DoVi? If so, do you still have the DoVi config file that can be shared?
@SamuriHL, have you? :) I think you have all the tools for this, haven't you? :) It would be really nice from you if you can give it a chance ... and we finally mange to get rid of the elevated black crap :)

SamuriHL

7th March 2021, 21:51

Hell no. These LG panels, especially the 18 series, are way too finicky to calibrate DoVi. As soon as you mess with the config it screws everything up in my experience. I've left HDR and DoVi alone other than SM white point @d65.

RXP

10th March 2021, 12:15

About due for an update here.

I spent way too much time testing all this curve stuff.
I've updated my settings for 130: https://postimg.cc/RJzNZd33
Pretty much just disabled FALL detail, it was too sensitive. FALL mod at 80% seems to be decent for extra brightness without losing much.

Curve is r13 now, probably my preferred for now. Hopefully the others like it too.
Not too sure how much the new normalization affects brightness, I haven't compared previous versions.

I'd reenable contrast recovery but I might wait, because I don't really seem to need it (personally, at least) with FALL mod.

Thanks for posting your curve/settings. I'm dipping my toe into the new builds and yours is the first time I've actually liked what MadVR has done on my C6 OLED. I just amended the peak brightness to 650 nits of my panel.

Edit: just realised you did 870 to stop the internal tone mapping right?

SamuriHL

10th March 2021, 13:14

It doesn't stop it but it is to choose the best curve the TV offers internally. It maps very close to eotf on our c8. And prevents the TV from clipping. You may have to experiment a bit with the c6 to see if it responds the same or needs a lower target nit.

Sent from my SM-G998U1 using Tapatalk

RXP

10th March 2021, 14:39

Excellent thanks! I just tried it on my C9 and on the jazz demo it did wonders! Even though I really like the C9's HDR performance, nice to squeeze that extra out of it.

quietvoid

10th March 2021, 15:30

chros

10th March 2021, 16:41

Thanks for keeping the spirit alive! :) I'll take a look when we will have a useable build (e. g. with an option for brightness normalisation).

SamuriHL

10th March 2021, 18:42

Excellent thanks! I just tried it on my C9 and on the jazz demo it did wonders! Even though I really like the C9's HDR performance, nice to squeeze that extra out of it.

The C9 and newer give you a few more options if you calibrate the panel. Because with those newer models, you can actually measure your peak luminance and create a curve to match on the panel. Then if you set madvr to tonemap to that measured peak, it should follow eotf to that point and wring every ounce of luminance you can drive out of that panel.

SamuriHL

10th March 2021, 18:43

So far the R13 curves remain my favorite, they are however extremely close to Fer15's new "low consistent" for everything below 10 000 nits.
R13 is still brighter in places and the difference is noticeable in HDR.

There's very little difference in highlight detail preserved, and the only I've found were for low nits (509 curves) at 130 DPL. R13 still has better contrast although it darkens shadows a little.

At this point I'm just trying to hypertune it as much as I can, and waiting on Fer15's opinion about it.

I like R13 a lot myself. I've not tested Fer's curves yet. I will be watching a movie tonight so perhaps I'll throw it on there and see what it looks like. Did you modify it for 870 nits by chance?

EDIT:

Just saw r15 and grabbed that. That's what I'll be using tonight. Thanks!

SamuriHL

10th March 2021, 18:44

Thanks for keeping the spirit alive! :) I'll take a look when we will have a useable build (e. g. with an option for brightness normalisation).

I know everyone keeps complaining about brightness normalization but everyone seems to fail to understand WHY he put it in there. It's so that people don't have to try to figure out brightness matching to compare curves. That option is not likely to be permanent once the curve(s) are selected.

quietvoid

10th March 2021, 20:06

There isn't a huge difference between R13 and R15 FWIW, it's mostly gonna be a difference for 10 000 nits and to retain contrast better instead of being flat.

SamuriHL

10th March 2021, 23:19

Well, we're at that point and that's where we want to be. :) That's seriously a good thing.

RXP

11th March 2021, 15:16

quietvoid

11th March 2021, 17:05

quietvoid, I did some more testing on my C6 with the default settings you gave. With the Spears and Munsil benchmark disk it's night and day! The detail increase is huge. Now most movies have very conservative HDR grades but good to know it's there for when they don't! It's a bigger difference than I got from doing a 3DLUT in SDR, which I thought was a game changer. For anyone that's on the fence like I was about MadVR with OLED's give it a shot, you'll be thankful you did.

Compared to the TV's tonemapping, it's definitely an improvement.

I've switched to my R16 (see signature) and changed my settings a bit: https://postimg.cc/1n14s6NB

I enabled contrast recovery because the improvement is significant even at the lowest strength.
There's slightly more blown out small highlights but the brightness is increased, and I most likely wouldn't notice the difference in highlights.

Bumped up DTN and lowered FALL highlight mod.

I'm still really conflicted with the desat lum method.
sep improves brightness and contrast significantly in S&M but I also absolutely can't stand how fires lose detail and saturation.
However it might end up being the leftover option since everyone uses it.

Hopefully the denormalization doesn't require too many changes.

SamuriHL

11th March 2021, 17:53

I was noticing the fire issue last night, as well. This may be something madman can address down the road when we get back to that. Interesting changes you've made to the settings, though. I'll check them out later, as well.

quietvoid

13th March 2021, 06:11

Another day another curve..
R17 is my most accurate curve yet, since it doesn't alter shadows much if at all.
I've compared against clipping and most of the time R16 has slightly brighter shadows, reducing contrast.

R17 is also slightly better with highlights, as long as other FALL/DTN options don't blow them out.
All this isn't very noticeable at high nits and HDR, but it's good to know that it's tracking PQ correctly.

Tweaked the settings a bit just because, I guess: https://postimg.cc/303ZkP1w

RXP

13th March 2021, 06:56

Thanks quietvoid, really excited to watch my first full HDR movie with your settings! - out of interest are you using v130 for watching content on your OLED or just testing? I flipped my OLED watching machine to v128 due to brightness norm and my PJ is on v130 as it's only 95nits peak

quietvoid

13th March 2021, 07:15

I'm just testing. I haven't watched anything in months.

chros

13th March 2021, 09:21

I'm just testing. I haven't watched anything in months.
:D Not good, not good :)

quietvoid

13th March 2021, 22:24

True.. it's even worse considering most content out there doesn't get any improvement from madVR.

quietvoid

14th March 2021, 06:45

So another curve that's mostly more for high nits, because it's even brighter for lower nits and I doubt people over at AVS are liking these much, even though they're accurate.
My idea is that a lot are used to darker video caused by having to rely on DTN for detail.

R18 is basically a brighter R17, without affecting highlights.
Since it's literally free brightness, I've gone to lower DTN and enabled max again. Settings: https://postimg.cc/c6Xs9dMT
The lower DTN is because max does reduce brightness in saturated areas, so it counteracts that loss.

max is noticeably better at highlight detail than sep, even though some specular highlights are dimmer because no desat.
I still think the significantly better fires and slightly more detailed highlights are worth losing a little contrast.

SamuriHL

14th March 2021, 15:36

I would agree with that assertion only because we have the room for it at 870 nits. Those who are trying to compress everything to a target below 100 nits have WAY less headroom to play with and thus have to make larger compromises. I'm good with losing a little bit of contrast to retain better highlights and better fire. I see you nuked FALL in your latest settings. Interesting.

quietvoid

14th March 2021, 16:13

Yea, it's not as easy for lower nits.
When tracking 1:1 with less brightness, the result is just flat because highlights aren't as bright.

I still think it's not that bad at 130 nits but not sure how it would look at 50 nits.
There's still a huge contrast increase for 2230+ nits because the original rendering intent is high contrast HDR, which 1:1 helps preserve.

As for FALL, since there's more brightness there's not much brightness difference between DTN 80 + FALL 90% vs DTN 75 + max.

I've just noticed that the 10 000 nits demo becomes flatter in some places with max because it's not desaturating, while sep keeps some contrasty bits in there.
Hard to tell how it's supposed to look, though.

SamuriHL

14th March 2021, 17:21

Compare it to the 1000 nit version. :)

quietvoid

14th March 2021, 17:48

I'm talking about the LG demo here, but it might be useful to compare S&M as well, I guess.

SamuriHL

14th March 2021, 18:14

Ah, ok. But yea, the SM demo is nice cause we have so many ranges of NIT values to compare the same content at. It can be useful.

Chouonsoku

15th March 2021, 07:09

I promise I searched before asking this but the answers I saw were a year old and I'm not sure if I'm interpreting the answers correctly. With an OLED that isn't calibrated with a 3DLUT, what is the correct setting in the calibration section? Using madVR b130 with quietvoid's settings and R18 curves. I've been using the following with the color gamut setting on my TV set to Auto and madVR outputting as HDR:

https://i.imgur.com/vjEUiKO.png

Should this be set to DCI-P3 with my TV set to Wide color gamut? And are these settings fine for SDR and HDR content?

quietvoid

15th March 2021, 14:45

DCI-P3 in madVR, TV set to Auto AFAIK.

Asked the same not long ago: https://forum.doom9.org/showthread.php?p=1933960#post1933960

I don't think there's a difference with using BT.2020, though..

VBB

15th March 2021, 19:17

It depends on which HDMI mode you use. For normal mode, you want BT.709 and the TV set to AUTO. For PC mode, because you are locked into WIDE, you want DCI-P3. The settings are only for SDR or HDR to SDR tone-mapped content.

Chouonsoku

15th March 2021, 20:16

Ah I gotcha. I don't use PC mode cus of the banding so I'll leave everything where it is. Thanks you two!

VBB

15th March 2021, 21:16

Does the gamma on your TV match madVR? I suggest you change that to 2.4, especially if you watch in a dark environment.

Chouonsoku

16th March 2021, 01:09

Does the gamma on your TV match madVR? I suggest you change that to 2.4, especially if you watch in a dark environment.

I have my SDR display modes set to BT.1886.

Asmodian

16th March 2021, 01:17

BT.1886 = 2.4 on OLEDs.

chros

16th March 2021, 08:06

Not on LG: bt1886 is slightly brighter near black than g2.4

VBB

16th March 2021, 18:16

@chros - Hmmm, that's not the case with my C7. I've done many back and forth comparisons with this pattern: https://www.avsforum.com/threads/lg-oleds-how-to-enhance-near-black-detail.2579537/post-46798617

Brightness has always been at 51, which is where 0% is truly black on my TV.

Chouonsoku

16th March 2021, 19:31

BT.1886 = 2.4 on OLEDs.If that's the case then why are both offered as Gamma options? There must be some difference. I read a long time ago that BT.1886 was better but that was when I first got my C7.

@chros - Hmmm, that's not the case with my C7. I've done many back and forth comparisons with this pattern: https://www.avsforum.com/threads/lg-oleds-how-to-enhance-near-black-detail.2579537/post-46798617

Brightness has always been at 51, which is where 0% is truly black on my TV.I thought any brightness setting above 50 began to raise black levels? Also, you're saying I should set both madVR and my TV to 2.4/BT.1886 gamma? I didn't think you were supposed to modify the setting in madVR unless your display is *actually* calibrated which mine is not, just standard settings adjustments but no white balance or anything like that.

Also, I feel like I derailed this thread a bit my questions so I apologize. :scared:

SamuriHL

16th March 2021, 19:38

Not derailing the thread at all.

As for bt.1886, it's not as simple as "= 2.4".

https://kb.portrait.com/help/bt-1886-10-questions-10-answers

It'll approximate it, but it IS different.

VBB

16th March 2021, 19:42

With perfect black level, there is no difference mathematically between 2.4 and BT.1886.

Black level depends on the panel. Since you have a C7, like I do, you could try 51 and see if there's any glow. You need to do this in a completely dark room and with a 0% pattern. My C7 starts to glow at 53. With 52, there is a very faint dither pattern if you put your nose on the screen ;) So I settled on 51.

I would set both the TV and madVR to 2.4 gamma. Calibrated in this case simply means match whatever you set the panel to. Doesn't matter if it's professionally calibrated or eyeballed.

I for one welcome the derailing. No offense to the guys testing the curves ;) Much appreciated!

SamuriHL

16th March 2021, 20:52

Hey the curve work is important! :P LOL Yes, ALL of these topics are relevant to our OLEDs when using madvr. I calibrated my SDR dark mode for rec709 2.4 gamma 100 nits. So my gamma setting stays at 2.2 on the C8 but the LUT is calibrated for 2.4. That gets a little confusing for people. Trying to calibrate my bright mode for DCI P3 2.2 and having some issues but it looks like I screwed up the calibration process by missing a step. Once that's done I'll set up profiles in madvr to be able to switch between my dark mode 2.4 rec709 and bright mode 2.2 dci p3. No need to limit one's self to only one setting. :D

Asmodian

16th March 2021, 21:35

If that's the case then why are both offered as Gamma options? There must be some difference. I read a long time ago that BT.1886 was better but that was when I first got my C7.

Swap between the two when viewing different scenes. Do you notice a difference?

On my C7, C9, and CX I did/do not.

Anyway, BT.1886 is at least very close to a gamma of 2.4, so 2.4 is a the correct setting in madVR.

quietvoid

17th March 2021, 04:41

Hmm, I just noticed Dolby Vision was locked into Wide Color Gamut.

chros

17th March 2021, 07:44

Yes, but I think it is not the regular wide mode, just as brightness/contrast are not the same in DV mode.

quietvoid

17th March 2021, 19:15

Looked at HDR10 again and I still want to believe. Wide looks closer to what I'd expect than Auto, with madVR set to DCI-P3.
But apparently it's still wrong.

Eh, switching from Dolby Vision to Auto looks about the same saturation levels, except madVR is even more saturated.
And madVR still has issues with darker shadows it seems, maybe because of the DTN.

SamuriHL

17th March 2021, 22:30

Very possibly. I had to factory reset the C8 for some stupid things it was doing with my LUTS. Now after the reset, for HDR it has some weird slider thing for color temp. Um, WHAT? WTF is that? I need to get it back to my D65 measured color temp.

VBB

17th March 2021, 23:06

LOL I just read your chat with Leon. Hope you get it all sorted out!

SamuriHL

17th March 2021, 23:14

Yea, stupid C8 wasn't resetting that LUT no matter what I did until I factory kicked it in the ass. As I said, cannon to kill a fly but it worked. I now have a dark mode rec709 2.4 LUT and a bright mode DCI P3 2.2 LUT. It's looking quite good. madvr will be happy about the DCI P3 LUT.

quietvoid

18th March 2021, 14:50

Was playing around with low DTN and it certainly looks closer to the Dolby Vision image, although less contrasty looking.

Now I don't know which I like better :(
Low DTN definitely seems to have the correct shadows, but everything is bright and specular highlights blend in with the surroundings.

At least now that the curves are better, the possibility is there.

SamuriHL

18th March 2021, 15:17

DTN is tunable which is great. The curves are looking REALLY good at this point. Now we have options.

Cinemancave

21st March 2021, 12:59

I've been using MadVR DTM with my projector for a long time, but want to try this out on my OLED now. I just want to know what settings you guys are using on the TV side of things and what format you are outputting this in?

I am having a hard time grasping this concept for the TVs when compared to the tonemapping for projectors: Converting HDR to a fixed SDR gamma curve seems somewhat "stable", but on an OLED we want to force it into HDR mode right? Which leads to MadVR DTM followed by the TV's own tonemapping on top of this? Or can you disable the TVs own tonemapping somehow - maybe putting it in HDR Game Mode and putting it into HGIG? If we are dealing with the TVs own DTM on top of MadVRs then isn't this impossible to figure out since we don't know what the TV is doing?

What would be some basic settings I should use to try this out?

SamuriHL

21st March 2021, 13:44

Look back to quietvoids latest post that contains a link to his settings and his signature has a link to his curve.

As for how this works a lot of testing went into figuring out what the TV does at various nit levels. At 1000 nits it more or less tracks eotf. That is what we want. So what we're doing is determining the real display nit for the oled and telling madvr to tonemap to that. This forces the TV to use an internal tonemapping curve that is less aggressive than the higher nit curves. You're still getting the TV tonemapping on top of what madvr is doing but because you're forcing it to use a curve that tracks eotf it's the least manipulation we can get. On later models of LG oled you can even calibrate your curve to match exactly up to your measured max nit level which is ideal.

Sent from my SM-G998U1 using Tapatalk

Cinemancave

21st March 2021, 14:12

Great thanks!

So in "Edit custom TM curves" I paste all curves from his signature, or do I choose one?

I have a CX and calibration equipment, has anyone written a how-to on matching up the curves?

quietvoid

21st March 2021, 17:31

You have to import the curves with the button at the bottom of the curve editor.
Not sure what you mean about matching the curves.

SamuriHL

21st March 2021, 17:46

It's not about "matching up the curves". I think my explanation confused you as I'm talking about 2 very different things. Let me try again.

Topic 1: madvr

This is using quietvoid's settings, and importing his custom curve into madvr while setting your real display nits to whatever you measure with your calibration equipment.

Topic 2: TV

This is where you can, if you have the right equipment and software, calibrate your TV's HDR mode to follow a curve to your display's real display nits.

https://www.youtube.com/watch?v=AJ_CYgfwo7o

Watch that video from Vincent for a better explanation.

SamuriHL

21st March 2021, 17:56

You have to import the curves with the button at the bottom of the curve editor.
Not sure what you mean about matching the curves.

I think I confused it by talking about matching the TV's tonemapping to follow EOTF using a custom curve. Which is true but it already does that up to 700 nits, as we know. On the C9 and newer OLEDs they can change that curve to follow EOTF up to the measured max luminance of the display so that it rolls off or clips at a higher point than the built in 700 nits tone mapping. In theory, this should allow them to mostly bypass internal tone mapping altogether if they're using madvr because the content will already fit within the TV's custom curve. Since I don't have a C9 or higher, I've no idea if that is true or not. Theoretically it should be, but, in practice with these panels? Who knows....

Cinemancave

21st March 2021, 18:15

Interesting how this all lines up: I did som googling and found the template for Device Control which allows the user to control the luminance point and tonemapping settings like Vincent showed in the video you just linked to. So I am going to use it and see how this all looks with 100% roll-off, basically clipping everything above my panel's peak brightness and then letting MadVR do its' thing :)

Thank you for your quick replies. I will report back.

VBB

23rd March 2021, 22:13

Since we're always wondering if Sony OLEDs behave differently: I just found out that the OP, Sunset1982 (a fellow German), actually owns an A9F. He is also planning to buy the new A90J. Last time I spoke with him over at AVS, he wasn't even aware he had created this thread. I'll try to get him to post his impressions here.

chros

24th March 2021, 07:10

That would be really nice, thanks, especially in graphic/game modes to get chroma 444 :)

Sunset1982

24th March 2021, 07:55

Hey guys, VBB told me I could help here out with my sony tv. I don't have a htpc atm, but maybe I can help you guys. I laso have my old LG C8 in my living room. I am very busy at work these days, but I will do what I can to read the last few pages of this thread. Any questions I can answer in the meanwhile or something to test for the weekend?

chros

24th March 2021, 09:37

Cheers, but if you don't have a PC with it ... :) I think we would be mostly interested in how it performs with chroma 444 (graphic/game modes) at 23p with SDR/HDR10.

Sunset1982

24th March 2021, 16:05

I still have a pc at home, but it is in another room. I could take it and test some things, but before that I have to know what exactly I should test and how (what patterns)

SamuriHL

24th March 2021, 16:26

They're really interested to see if 444 mode introduces artifacts in the video. Banding is one of the biggest issues. Near black level issues are another. I'm sure they'll dig up the patterns again for you to try. It would be useful indeed. Especially given that these consumer OLED panels are all made by LG with different electronics based on the MFG of the TV itself. Has Sony solved these issues? Or are they still prevalent.....that's what they wanna know.

chros

25th March 2021, 01:40

Yes. But first of all can you tell us for sure whether it has the new Evo panel? :) (It's still an impressive set even if it doesn't.)

SamuriHL

27th March 2021, 22:57

Leon is the best. So my C8 I had a mode calibrated for rec.709 2.4 dark and another DCI-P3 2.2 for bright. But I learned today from Leon that these LG OLEDs have 2 different LUT slots on each picture mode....one for SDR rec.2020 and another for SDR rec.709. And with the latest device control and LG templates for ColourSpace, you can load them both at the same time. :D So for the moment I just did my bright mode to have DCI-P3 for the SDR rec.2020 LUT and rec.709 for the rec.709 LUT. To get the LG to switch between them you simply need to force the bt.2020 flag. My TiVo already does this when outputting 4k. And the SHIELD can do that if you tell it to always output bt.2020. And as we know, madvr can be told to send the bt.2020 flag on nVidia cards. It works beautifully!

chros

28th March 2021, 01:35

Yes, indeed, that's how hdr10 works with the bt2020 slot, and that's why DoVi uses the bt790 slot (so it doesn't trigger LG's sh!tty gamut mapping)! :)
The problem with this is the aformentioned low APL issues (e.g. cyan/magenta) where LG's gamut mapping failes miserably.
Why don't you upload the dcip3 into the bt709 slot? (I haven't tried it personally.)
And you can do the same with the python tool.

SamuriHL

28th March 2021, 03:02

I did that and rec.709 content becomes oversaturated. By having a rec.709 LUT in the rec.709 slot and a DCI-P3 LUT in the bt.2020 slot, when the bt.2020 flag is enabled it uses the DCI-P3 LUT and when the bt.2020 flag is not enabled, it uses the rec.709 LUT. With this setup, all content looks damn near flawless. The nice part about ColourSpace is that I was able to generate everything from one calibration session's data just by generating the LUT that I wanted. And indeed, I looked at some upscaled 1080p->4k videos using DCI-P3 with the bt.2020 flag enabled in madvr and man it's REALLY nice. I'm quite impressed with this setup.

quietvoid

1st April 2021, 01:51

Every time I read about calibration on these TVs I get overwhelmed with info and no idea where to even begin, as there doesn't seem to be a concise guide on manual calibration.
It's even worse for HDR :)

Asmodian

1st April 2021, 02:46

What do you mean by manual calibration?

SamuriHL

1st April 2021, 02:53

Don't even waste your time trying HDR and DV calibration on these panels. They aren't stable enough. Calibrate the white point balance in the service menu and call it a day for those. For SDR calibration, I use colourspace to create a 3D lut and upload it to one of the picture modes (bright mode, dark mode, whatever). I got a LOT of help from Leon and Ted when I first started out with ColourSpace but it's become much easier to use with each new beta. The quality I get from SDR on the C8 after calibration is nothing short of amazing.

But for HDR, don't bother. Lots of people do, of course, and they're welcome to. But it's not worth the effort.

quietvoid

1st April 2021, 04:01

What do you mean by manual calibration?

I think it would be everything other than AutoCal, but that might still be possible for SDR.
I think the issue with HDR is that more expensive equipment or manual intervention is necessary for the 2018 models, while newer have an internal pattern generator?
Though I've noticed most people do HDR calibration with other software than Calman too.

Don't even waste your time trying HDR and DV calibration on these panels. They aren't stable enough. Calibrate the white point balance in the service menu and call it a day for those. For SDR calibration, I use colourspace to create a 3D lut and upload it to one of the picture modes (bright mode, dark mode, whatever). I got a LOT of help from Leon and Ted when I first started out with ColourSpace but it's become much easier to use with each new beta. The quality I get from SDR on the C8 after calibration is nothing short of amazing.

But for HDR, don't bother. Lots of people do, of course, and they're welcome to. But it's not worth the effort.

To be honest I just notice issues in HDR, I have a hard time with natural skin tones in HDR10 only, not even DV.

Maybe there could be improvements in SDR but so far it's been good enough.

I guess I could be looking at this: https://www.lightillusion.com/manual_calibration_guide.html
But even then Calman seems to be cheaper..

SamuriHL

1st April 2021, 13:01

I really don't want to get into a comparison between calman and other software and why I dropped calman like a bad habit. Cheaper, however, is a point of view. If you like yearly license fees, then sure calman with Autocal is cheaper... for a couple years anyway.

As for hdr calibration on these panels, it's not about the pattern generator. Pgenerator is an excellent bit accurate external pattern generator and with an hdfury you can do hdr with it. You can also use madtpg if you take the time to set things up properly.

The problem is these panels aren't stable enough to calibrate with hdr. And you can't generate a full 3D lut for hdr. You get a matrix lut which I think is doing primary colors for the most part. The one thing I've seen make a real difference is calibrating the white balance in the service menu. You take one of the temperature settings and calibrate that for d65. Then you set your picture mode to use the color temp that you calibrated. That has a real impact and is about the best you can realistically do for these panels.

I stopped chasing the hdr dv calibration dream a year ago for these panels. I've not looked back.

Sent from my SM-G998U1 using Tapatalk

quietvoid

1st April 2021, 13:54

Oh well. Not sure I'm willing to spend 300 USD just to calibrate white point :D
Thanks for the info.

SamuriHL

1st April 2021, 14:10

There are other tools where you could do that much cheaper.

https://sourceforge.net/projects/hcfr/

quietvoid

1st April 2021, 14:25

Yes, but I don't have a colorimeter :)

SamuriHL

1st April 2021, 14:32

Well that would certainly hinder any efforts for calibration, yes. :D

chros

1st April 2021, 16:40

@SamuriHL, which TPG did you use?

Yes, but I don't have a colorimeter :)
:D Well, you definitely need one for any calibration/profiling software.
You can get even a used id3pro, or just wait for a price cut e.g. on amazon.

Funny you asked this, we just playing around with this in the last 2 weeks :)
Personally I don't care about SDR either, you can create a decent 3dlut with DisplayCal (either for madvr or for the TV, I only use madvr for SDR).

But about the other 2 :), I wanted to see whether:
- in HDR10 the bloody cyan/magenta issue can be dealt with
- in DoVi removing the glowing in near black scenes (the elevated black + black crush can't be fixed with 1dlut/calibration)

You can read about the process from here (https://hub.displaycal.net/forums/reply/29232/), it's essentially 1dlut calibration (and with synthetic 3dlut, if it's needed), you need HCFR and aiopylgtv (https://github.com/bendavid/aiopylgtv) for this.
I used madTPG as TPG, but even setting this up is problematic but only when using HDR10 output!:
- using the old driver (see my signature) that keeps the display in HDR even when app is not fullscreen
- run a calibration pattern video with metadata 1000/400 (maxCLL/maxFALL) and P3 primaries that triggers the display into HDR10 mode and stays in it
- run madTPG for SDR calibration! :)

I tried tons on different combinations, the results are questionable at best :D

1. HDR10 (in calibration mode, aiming gamma 2.2 for 1dlut):
- Calman: I was interested what it does and can it get rid of our issues. In short: it can't, the issue is still there :D
Also, due to the generated 3dlut (if it's used), colors are completely different from the reference cinema preset :D
- manual 20 point grayscale calibration (TV controls + HCFR): the result is way brighter than the reference preset (I guess due to the used factory 3dlut), our issue is still there :)
- manual 1dlut (HCFR + aiopylgtv): with this we have multiple options: with factory 3dlut or synthetic 3dluts:
- - with factory or absolute synthetic 3dluts: the issue is still presnt :)
- - with perceptual / preserve saturation / saturation synthetic 3dluts: the issue is gone!!! But obviously colors will be off, but after actually seeing what Calman does ... :)

So, now I have 1 preset with custom 20 point 1dlut + saturation synthetic 3dlut, the closest we can get to "reference" preset without having the bloody issue!

Interesting thing is that C9 still has this, C10 users couldn't verify this, but Vincent pointed out some color related issues in his newer DV vs HDR10 video, Philips users also complain about something similar, but interestingly for them it happens with different scenes! :) So it seems that it's a problem with the given brand's implementation.
But why DoVi doesn't have this issue, what do they do differently?! Only if we knew ... :)

2. DoVi, it's more problematic, because you would need external devices to trigger it into Relative mode (to measure gamma 2.2) that I don't have:
- it turned out upon experimentation, that Game preset is just a bit brighter than Cinema and it doesn't have the glowing near black! (neither Cinema home, but that is way brighter)
- manual "blindly" adjusted 5-20 point didn't get rid of the glowing
- I uploaded 700 nits 1dlut (+ factory 3dlut) into Game preset that resulted in / compared to:
- - Cinema (dark): brighter image in normal and bright scenes, but same lightness in dark scenes
- - Cinema (home): darker image in dark scenes, around the same lightness in normal and bright scenes
- - both Cinema and Cinema Home have more vivid / lifeful colors, the 1dlut one is a bit duller colorwise

So, in summary, the image is different than the factory presets and not in a good way. Also, it seem DoVi mode uses different gamma / or-something, otherwise the image wouldn't be brighter for the dark preset.
So, back to factory luts, it seems no dice with DovI for us, only LG can fix their sh!t :)

SamuriHL

1st April 2021, 17:35

I tried with both pgenerator (rpi) and madtpg. The problem is there is more than one issue. You can't generate a stable 3D lut for HDR or DV as you know. But, worse, is the bug in the LG panels (ALL of them afaik) that if you upload a 1D lut you've essentially f***ed up black levels. So it's kind of a nonstarter. Would be nice to be able to do the 1D lut to correct grey scale but as soon as you do it black levels are borked due to the LG issue. This is why with ColourSpace we avoid the 1D lut like the plague and focus only on 3D lut which for HDR/DV is pointless. So, yea...factory luts indeed.

quietvoid

1st April 2021, 17:39

I'm curious if this is the same issue as what you refer to as the cyan/magenta issue:

HDR10 / Dolby Vision
https://i.postimg.cc/kBvMHf6G/HDR10.jpg (https://postimg.cc/kBvMHf6G) https://i.postimg.cc/G433nGGQ/DoVi.jpg (https://postimg.cc/G433nGGQ)

Unfortunately if I were to buy a colorimeter, I could only make use of it for the OLED TV, since my current Linux setup has no support for ICC profiles either.
Hard to justify the cost given that I've only got ~670 hours on the TV in the past 2 years.

SamuriHL

1st April 2021, 17:40

P.S. I have the equipment to do a DV calibration and have done so before. The results were a mixed bag. Some things did improve. Others, like black levels, were much worse. I came to the conclusion that everyone else that is far better at this than me came to....Don't bother and just do the SM WB correction and leave the factory luts alone. It's annoying cause if LG fixed their s*** as you say, then we could definitely improve things especially with greyscale. But....no.

QBhd

2nd April 2021, 04:04

.... I've only got ~670 hours on the TV in the past 2 years.

LOL

In less than 2 years (I just checked, it is 2 days short of my 2yr anniversary!!) I have 4,545 hours

QB

chros

2nd April 2021, 09:45

I'm curious if this is the same issue as what you refer to as the cyan/magenta issue:
I don't think so, the issue affects low APL scenes, if you have netflix DV then you can see it for yourself with this sample (https://forum.doom9.org/showthread.php?p=1925528#post1925528).

About luts, if you want, I can upload 1dlut and 3dlut to try out if you have a spare preset (e.g. technicolor).

I tried with both pgenerator (rpi) and madtpg. The problem is there is more than one issue.
So which one did you use finally? I'm asking because as it turned out ColorSpace has an issue with madvr (former wokrs with full range, latter requires limited).

This is why with ColourSpace we avoid the 1D lut like the plague and focus only on 3D lut which for HDR/DV is pointless. So, yea...factory luts indeed.
This works with SDR, but who cares about it anymore?! :)

But, worse, is the bug in the LG panels (ALL of them afaik) that if you upload a 1D lut you've essentially f***ed up black levels. So it's kind of a nonstarter. Would be nice to be able to do the 1D lut to correct grey scale but as soon as you do it black levels are borked due to the LG issue.
Do they actually know what they talk about? Couple of measurements in calibration mode:
- cinema HDR with factory luts (https://imgbox.com/g/prESCcCuKa)
- cinema HDR with unity luts (https://imgbox.com/feP2kNRj#)
- cinema DoVi with factory luts (https://imgbox.com/M90rDRkN) (Oppo displays HDR10 20 points HCFR grayscale patterns and upconverts them into DoVi)

Now, if you flatten the gamma to power gamma 2.2, then you will get brighter image!
While experimenting with synthetic 3dluts, it's clear that there are bunch of possibilities that will modify the overall gamma no matter what 1dlut is used (so you have to modify the 1dlut while the 3dlut is in place), and we don't have a clue how the factory absolute 3dlut was created, let alone how actually the whole chain is supposed to work (not just in theory).
My best guess now is that the gamma shouldn't be flat 2.2, but the rotated S curve we have seen in those images.

SamuriHL

2nd April 2021, 12:09

I use pgenerator exclusively now.

I've not tried playing around with synthetic luts. I've seen people post about them but creating them is above my pay grade. I'm by no means any kind of expert in calibration at all. I know the process that was taught to me but deviating from that is beyond my capability.

As for who cares about sdr well I still do as I watch a lot of upscaled content. Having an accurate sdr picture mode is nice. Maybe someday we'll get a TV that allows us to have accurate hdr and dv too.

Sent from my SM-G998U1 using Tapatalk

quietvoid

2nd April 2021, 14:05

I don't think so, the issue affects low APL scenes, if you have netflix DV then you can see it for yourself with this sample (https://forum.doom9.org/showthread.php?p=1925528#post1925528).

About luts, if you want, I can upload 1dlut and 3dlut to try out if you have a spare preset (e.g. technicolor).

Oh okay, no idea what's happening with my HDR10 then.
It seems to just have really not smooth shadow to midtone gradients.

Dynamic tonemapping helps brighten things and looks better.
Same for dynamic contrast.

Otherwise I couldn't find other settings that improve anything.

Wonder if I can generate a 3D LUT myself with DisplayCAL without any equipment?
Otherwise it could be useful to try out yours.

Couldn't exactly reproduce the config discussed over at the DisplayCAL forums, and attachments require an account :|
Also seems that I can't create a 3D LUT, button is greyed out. Edit: Okay there's a separate LUT creator.

SamuriHL

2nd April 2021, 15:12

Generating a LUT without measuring is....um, less than useful. That's because natively every panel is different. When a pixel asks for red (255,0,0) or (235,0,0) your native panel will display what it thinks of as red. But there's going to be a deviation error. Maybe it's small under 5. If so, great. Maybe it's bigger. The way to determine that is to measure it against a calibrated meter. The more accurate the meter, the better, which is why professionals profile the meter. I don't have a way to do that with mine so I just use my meter. Anyway, using all those measurements, the LookUp Table (LUT) is created to correct your native panel's deviations. Maybe red is really (254,1,0) or (234,1,0) or whatever. The LUT tells the panel how to correct it. I get it, this is all basic knowledge and all that. BUT the reason I mention it is because a LUT created for one panel is going to very likely wildly throw off another one and make things worse not better. Creating a "generic LUT" is not really possible. It must be measured to be accurate.

quietvoid

2nd April 2021, 15:20

I'm aware of all that, I'm just trying to figure out if my panel's default HDR LUT is causing issues.
So far I tried uploading a synthetic LUT but it was really bad (probably my fault).

Then I did a DDC reset and the image is much improved, but my white balance controls are now disabled for some reason and I can't fix it :(
Comparing my now borked technicolorExpert preset to Cinema, shadows seem much more accurate.
Before doing anything, both presets would look the same.

Maybe I'm stuck with having to do a factory reset.
Ah: WARNING: When running the ddc_reset or uploading LUT data on 2018 models the only way to restore the factory LUTs and behaviour for a given input mode is to do a factory reset of the TV.

Now I just don't know how to set Warm2 back.

quietvoid

2nd April 2021, 15:55

Anyways, factory reset and reuploaded the synthetic 3D LUT and my issue isn't fixed.
I don't know what the 1D LUT is for but it seems to be the culprit.

Edit: Apparently it's for the greyscale in HDR, so maybe I'm actually gonna have to get a colorimeter to correct this..
j82k's 1D LUT is brighter (no idea if correct) and fixes the issue but now I doubt my whitepoint is correct, seems too cool.

SamuriHL

2nd April 2021, 16:46

I suppose I can take another crack at HDR calibration at some point. I've not done one in ColourSpace as I don't typically touch 1D lut for anything, but, I suppose I can find a mode I'm not using (gaming mode perhaps) and try a calibration and see what happens. Won't be til later this weekend if I do it.

EDIT: Also something to be aware of. Uploading a 1D lut BYPASSES the service menu white point adjustments.

quietvoid

2nd April 2021, 17:01

Right, I guess the only way to correctly do the greyscale is with a proper (not synthetic) 3D LUT.
Otherwise I have no idea how to adjust whitepoint, maybe with both 1D and 3D LUTs like @chros is doing..

Still a bunch more research to do.

SamuriHL

2nd April 2021, 17:48

The greyscale is done through the 1D lut. The 3D lut handles the entire colorspace. Think of this as building off of the 1D lut. But LG has stupid bugs with 1D luts that cause all kinds of irritating issues. Supposedly the latest Calman beta "fixes" those issues or works around them or some such. I've no desire to shell out more money to find out. The reason the 1D lut bypasses the service menu white point, though, is that it contains the entire greyscale correction from white (255,255,255) 109% to black (16,16,16) 0% (assuming video levels not full). The idea is that you want to filter through as little processing as possible. I have my doubts about whether we can get a proper HDR calibration but hey I'm game to try. Also, just so you're aware, when doing the LUTs for HDR, you measure at 2.2 gamma and when the LUT is uploaded it's converted to PQ so I'm told. So it's not REALLY a 2.2 gamma for HDR even though the TV uses that to calculate it. I've no idea how that works in reality but that's what calman and colourspace tell us.

quietvoid

2nd April 2021, 21:37

Thanks.

I've taken the 1D LUT and set all points 1:1, that helped me confirm my factory HDR 1D LUT is garbage.
At least I hope that's what it means, because details in the shadow gradients are way better, and nothing else increases in brightness.

Now I can't fix the whitepoint without measurements.
Still unsure whether my panel is pushing red a lot or if my new 3D LUT is useless.

Debating whether I should "future proof" with the i1 Display Pro Plus (2000 nits) or not.
Everything is so expensive when it has to be imported in Canada..
Then there are the i1d3 OEM clones which are way cheaper as well, but probably older revisions.

Ugh.

dwaleke

2nd April 2021, 22:10

Go with the i1D3 OEM.

Same as the Pro Plus retail, but does not require special unlock codes to work with 3rd party software.

QBhd

2nd April 2021, 22:16

I bit the bullet and got i1 Display Pro Plus a while ago.... I haven't played around with it for a long time, but maybe it's time to do so again... you can get them from Vistek and Henry's

https://www.vistek.ca/store/435802/xrite-i1display-pro-plus

QB

quietvoid

2nd April 2021, 22:16

Go with the i1D3 OEM.

Same as the Pro Plus retail, but does not require special unlock codes to work with 3rd party software.

There aren't any OEM options here. Most retail devices are supported in DisplayCAL, HCFR anyways.
The only OEM I can find are from Calman directly, but it's even more expensive given that they don't include import fees, and charge 40$ for shipping.

Maybe I'll just wait for my local shop to have stock of the Plus again, for some reason it's cheaper than the regular Pro..

you can get them from Vistek and Henry's
They sell it 60$ more than my local store :D

Alternatively Amazon is ~20$ more (only one tax) and delivers this week..
Committed to the Amazon one.

kostik

2nd April 2021, 22:29

I also moved to ColourSpace. I wanted to know... Is it true that it is enough to calibrate Sony OLED SDR and it applies the calibration on Dolby Vision and HDR ??
A bit off topic .. I know

SamuriHL

2nd April 2021, 23:21

I don't know enough about Sony panels, but assuming they work similarly to LG here's what we do (and what I just did as I hadn't done this second part yet):

In our service menu, we calibrate one of our temperature modes to the white point we wish to use. I use D65 since any AWP will be pointless without a spectro to profile my I1D3.

Next, in the modes you want to use, change the color temp to the one you calibrated. E.G. I calibrated my cool temp and I set all my modes (SDR bright and dark, HDR Cinema and Technicolor, Dolby Vision Cinema Home and Cinema).

Then we calibrate SDR the way you want by creating a 3D LUT. My dark mode is rec.709 2.4. My bright mode has both DCI-P3 bt.2020 and rec.709 both at 2.2.

What I just did for the first time is used ColourSpace to "calibrate" my technicolor HDR mode. I put that in quotes for a reason. I did NOT mess with the LUTS here. Both 1D and 3D are factory still. But, having set the white point to D65, I used Leon's instructions to adjust the 2 point white point balance (high and low) to tweak the calibration slightly. Doing this, at least according to what my comparisons of a 22 point greyscale measurement sweep, looks like I have a tighter RGB separation both near black and near white. That should improve things noticeably in my HDR content.

So, back to your Sony, if it works similar to how our LG panels do, then if you can calibrate your white point temperature and then use that in your HDR and DV modes, then yes, you'll see a benefit.

chros

3rd April 2021, 10:05

j82k's 1D LUT is brighter (no idea if correct) and fixes the issue but now I doubt my whitepoint is correct, seems too cool.
He had a replacment panel back in the day that results in black crush, that's why he was interested in creating 1dlut.

I've taken the 1D LUT and set all points 1:1, that helped me confirm my factory HDR 1D LUT is garbage.
At least I hope that's what it means, because details in the shadow gradients are way better, and nothing else increases in brightness.

Now I can't fix the whitepoint without measurements.
Still unsure whether my panel is pushing red a lot or if my new 3D LUT is useless.
So this the settings (https://imgbox.com/g/vIx0qDD5Ae) we used creating synthetic 3dluts (1 of them is bt2020, the other is dci-p3). Note that every setting you modify, the outcome can/will be different (modifies the resulting gamma).
LG and Calman also use absolute rendering intent.

Upload the desired 3dlut at first or leave the factory one in place, then try to create a 1dlut.
But it won't be 10 minutes job :) And you can easily introduce other artifacts that can be only visible if you switch between your and a "reference" preset. (You can set up Cinema Home to be almost identical to Cinema preset: color temp w50, gamma medium, gamut Extended).
Then when you think you're done, check the result with various clips you are familiar with and switch between presets.

The biggest issue is that we don't have reference image to compare to, so no one really knows how certain scenes are supposed to look like.
E.g. I just noticed by accident that LG factory luts paint way more blue in low APL scenes, this can be seen with the factory measurements (https://imgbox.com/K1ZI6WyY) as well and can be fixed with 1dlut and can easily seen with this clip (https://we.tl/t-Jw3fvmUK1z) along with the cyan/magenta issue (file will be up for a week).

kostik

3rd April 2021, 12:18

SamuriHL

3rd April 2021, 13:17

That's entirely possible it works that way. These panels take the 2.2 gamma and covert it internally to pq eotf. That's why we calibrate greyscale at 2.2 for dv and hdr. I know for LG it's not that simple as just doing an sdr calibration. We have to calibrate the actual dv picture mode which means setting the proper dv Metadata while calibrating. Kind of a pain in the ass.

Sent from my SM-G998U1 using Tapatalk

quietvoid

3rd April 2021, 15:27

Upload the desired 3dlut at first or leave the factory one in place, then try to create a 1dlut.
This is the opposite of what I've read, usually it's 1d first, 3d after.

At this point I just need proper measurements to be able to know what's up with the panel :)

Though I still want to test what the whitepoint is for the HDR Cinema preset once 1D lut is uploaded, because I've been testing with technicolor Expert and that defaults to Warm1.
And I've been using weird profiles for the 3D LUT to adjust to a different WP, but the end result loses a lot of red saturation (compared to factory).

I want to believe I won't need to use a 1D LUT but won't know until I actually try to calibrate.
Not sure there's any other way to adjust luminance for HDR.

chros

3rd April 2021, 15:31

but its off topic so it doesn't matter...
It's not offtopic, shoot any OLED related questions, probably the thread needs to be renamed :)
@Sunset1982 knows more, he has a Sony now :) Btw, how is it?

SamuriHL

3rd April 2021, 16:16

You can't create a true 3d lut for hdr on these panels. That's why calman offers a matrix lut that does primaries. But you are correct that if you're going 1d lut for greyscale correction that it should be done first else the 3d lut becomes invalid. Alternate white point isn't something I've played with because I don't have a spectro to profile my meter therefore it won't be accurate enough. The 1d lut will bypass any service menu calibration you might do.

Sent from my SM-G998U1 using Tapatalk

quietvoid

3rd April 2021, 16:43

Welp, I don't see any other option for HDR if fixing the greyscale is necessary, it's either 3D LUT to get back to D65 or nothing.

SamuriHL

3rd April 2021, 17:08

Well, there's always the way I did it which does not involve messing with the factory LUTS. Fix white point to D65 in the service menu, and then perform a manual adjustment to the 2 point white balance for the picture mode you wanna use. That's a "good enough" calibration for HDR on these panels.

quietvoid

3rd April 2021, 17:25

Yea but it probably will not fix my shadow brightness, which is what requires the 1D LUT :)
I'm probably going to try just doing white balance first, and see if there's still a noticeable problem in the greyscale brightness levels.

chros

3rd April 2021, 17:26

Fix white point to D65 in the service menu
At which nits?! :D 100/192/203 nits? You get my point ...

then perform a manual adjustment to the 2 point white balance for the picture mode you wanna use. That's a "good enough" calibration for HDR on these panels.
According to those who didn't dare to touch the 1dlut ...

This is the opposite of what I've read, usually it's 1d first, 3d after.
Calman also does this way (with its workflow).
3dlut using absolute rendering intent shouldn't touch gamma, but due to whatever LG does with its bt2020 processing, it does even using absolute rendering intent (you can measure it and it's also visible).

Though I still want to test what the whitepoint is for the HDR Cinema preset once 1D lut is uploaded, because I've been testing with technicolor Expert and that defaults to Warm1.
This is just a setting, you can switch it to Warm2 as well.

I want to believe I won't need to use a 1D LUT but won't know until I actually try to calibrate. Not sure there's any other way to adjust luminance for HDR.
About this: once you get to the point to adjust grayscale (either with TV WB controls or 1dlut), make sure you aim at 700 nits (with 100% white), that's what the internal tonemapping requires, if you go higher (even just 730 nits) the whole picture will be brighter than reference (this can be also measured using PQ curve in HCFR).

chros

3rd April 2021, 17:31

I'm probably going to try just doing white balance first, and see if there's still a noticeable problem in the greyscale brightness levels.
The TV settings are supposed to be used in PQ (not in gamma) and unfortunately the lowest setting is C268 that can't go low enough to correct low APL issues (~10% in gamma). I tried ... :)
This is how they align with B8 approx. (some C8 values are slightly different) Also C1023 means 109% white not 100% :)

10% C268
15% C335
20% C402
25% C436
30% C470
35% C485
40% C500
C516
45% C531
C549
C563
55% C579
60% C594
65% C610
70% C627
75% C643
80% C658
85% C674
90% C692
109% C1023

quietvoid

3rd April 2021, 17:35

This is just a setting, you can switch it to Warm2 as well. Not after uploading the 1D LUT..
Which is why I'm stuck with a bad white point right now in HDR, with no idea how to correctly adjust it without measurements.

About this: once you get to the point to adjust grayscale (either with TV WB controls or 1dlut), make sure you aim at 700 nits (with 100% white), that's what the internal tonemapping requires, if you go higher (even just 730 nits) the whole picture will be brighter than reference (this can be also measured using PQ curve in HCFR).
The TV settings are supposed to be used in PQ (not in gamma) and unfortunately the lowest setting is C268 that can't go low enough to correct low APL issues (~10% in gamma). I tried ... :)

You mention TV WB controls for the grayscale, but don't those only adjust the color values, not the actual brightness tracking? At least in HDR that's the case.
In SDR you can control luminance for every 5% stimulus, but not in HDR.

Or I'm missing something?

What I'm trying to fix is the PQ tracking, which I'm suspicious about the TV tracking accurately in the low-mid levels.
Not the colors.

chros

3rd April 2021, 18:05

Not after uploading the 1D LUT..
Oh, you are right, but as @SamuriHL said, the uploaded 1dlut is overriding it anyway!
And @jk82 used an alternative white point (if you use his 1dlut), maybe he mentioned it in the displaycal thread which one exactly.

You mention TV WB controls for the grayscale, but don't those only adjust the color values, not the actual brightness tracking? At least in HDR that's the case.
In SDR you can control luminance for every 5% stimulus, but not in HDR.

Or I'm missing something?

What I'm trying to fix is the PQ tracking, which I'm suspicious about the TV tracking accurately in the low-mid levels.
Not the colors.
Exactly that's how you can modify luminance as well, by reducing/increasing e.g. all of them (RGB) at a given point(s). The side effect is that colors will be also affected, not just because of modifying the controls but altering gamma as well.

Edit: and that is exactly what the 1dlut also does. You have 4 values: input signal (from 0 to 1) and the corresponding 3 RGB values.
So essentially the 2 approaches are the same (if LG handles them equally), but you can set more data points (and wherever you want) with 1dlut.

Edit2: you can try this out even without a meter, add e.g. 20 to every RGB values at every point. :)

quietvoid

3rd April 2021, 18:31

Pretty sure I had tried playing with increasing all the RGB values but I don't think it made a difference in the TV settings. Gonna have to try again.

I'm not too sure about your interpretation of the PQ values, though.
C268 would be around 7-8 nits, is that not low enough?

I don't think my issues are that low brightness anyways, will have to try and see if there's a difference when I increase all RGB.

chros

3rd April 2021, 18:44

I'm not too sure about your interpretation of the PQ values, though.
C268 would be around 7-8 nits, is that not low enough?
Unfortunately not low enough to adjust near black (low APL) issues, but you will see.

I have uploaded for you couple of 1d/3dluts (https://we.tl/t-gJFCasiEAD) (file is up for a week) that you can play with technicolor preset:

You can NOT use the *.cube files (3dluts): it's with 17 points (B8)!!!
- you have to create yours with 33 points (C8): use the settings in the images that I shared before and only change the resolution and rendering intent:
- btdc-is-g22-10000.cube : saturation rendering intent (-is)
- btdc-iaw-g22-10000.cube : absolute with white point scaling rendering intent (-iaw)

There are 4 1dluts (*.cal files) that you can use (B8 requires 700nits at 100% as well), all set around D65 WP:
- chdr15-iaw-730nits.cal: 730 nits (@100%) for -iaw 3dlut (with unity 3dlut applied)
- chdr30-iaw-700nits.cal: 700 nits (@100%) for -iaw 3dlut (with unity 3dlut applied)
- chdr35-iaw-700nits-3dlucorrected.cal: 700 nits (@100%) for -iaw 3dlut (with custom -iaw 3dlut applied), that's when I tried to correct the cyan/magente issue, and introduced other issues with it :)
- chdr88-is.cal: 700 nits (@100%) for -is 3dlut (with custom -is 3dlut applied)

I'll be curious which one you'll prefer :)

SamuriHL

3rd April 2021, 19:12

At which nits?! :D 100/192/203 nits? You get my point ...

No, your point is confused. Because you're talking about what the content was mastered to, not what the display is calibrated to. Do you think LG calibrates to 100/192/203 nits from the factory? D65 in the service menu is calibrated to 100 nits.

According to those who didn't dare to touch the 1dlut ...

That's a VERY bold, incorrect assumption you're making here. You truly believe that Light Illusion didn't actually test this thoroughly before giving the advice to not touch the 1D LUT on LG displays? Umm, you can believe that if you want I guess...

SamuriHL

3rd April 2021, 19:17

The other problem is our consumer meter (i1d3) does not measure low light very well. So calibrating to fix those kinds of issues with this meter is somewhat pointless. I know some people have created synthetic luts to solve this issue. That's beyond my capability. And you have to be very careful not to introduce banding, posterization, and other fun and exciting artifacts with all of this. Exact science my ass! LOL

SamuriHL

3rd April 2021, 19:46

Just so we're clear about the 1D LUT issues I'm talking about:

https://liftgammagain.com/forum/index.php?threads/lg-oleds-3d-lut-profiling-using-lightspace-thread.13105/#post-131512

quietvoid

3rd April 2021, 21:11

Quoting a post from there:
So its better to not reset the 1D LUT table, do manual cal of 100% White using RGB-High and set Brightness to a value that is correct for the TV near black performance and then perform a 3D LUT, where inside to the 3D LUT it will contain 33-Point Grayscale also.

I guess this only applies to SDR, because HDR has no brightness control and messing with 20 points is sub optimal because artifacts.

I tried fixing my factory preset in HDR and it didn't work at all with with 2 or 20 points.
So I still want to believe resetting the 1D LUT properly (with accurate whitepoint correction) is gonna be better for me.

Edit: Isn't this also what Calman does? I'm not sure there is manual HDR/DV calibration, only measurements and straight into a proper 1D/3D LUT combo uploaded to the TV.
At least for anything requiring grayscale brightness adjustments, whitepoint only doesn't require any of this.

SamuriHL

3rd April 2021, 21:41

Yes, Calman does a unity 1D LUT and 3D LUT for SDR, HDR, and DV. They claim the beta version that's been released in the last week works around/fixes the raised black and other issues with uploading a 1D LUT. My license ran out last November so I can't try it to find out if that's true. But they set a unity 1D LUT and then calculate greyscale corrections in that 1D LUT. Then they run a matrix LUT (or lightning I don't know what it's called these days) to create a primary 3D LUT. That's if you use the autocal workflow, anyway.

chros

4th April 2021, 12:29

No, your point is confused. Because you're talking about what the content was mastered to, not what the display is calibrated to. Do you think LG calibrates to 100/192/203 nits from the factory? D65 in the service menu is calibrated to 100 nits.
Yes, for SDR. But we are talking about HDR, that's why I said that it's far from simple and maybe no one knows how it's supposed to work.

You truly believe that Light Illusion didn't actually test this thoroughly before giving the advice to not touch the 1D LUT on LG displays?
As far as I know it's not Light Illusion who provided those extra tools :) But do they offer 1dlut upload capability now? (They didn't a year ago.) That means you have to accept everything they say, you can't try it for yourself.
And there's a valid usecase for 1dluts: e.g. to fix black crush (caused by panel replacement/firmware/etc), again with HDR.

And you have to be very careful not to introduce banding, posterization, and other fun and exciting artifacts with all of this. Exact science my ass! LOL
That's absolutely true :) I've created around 10 1dluts for now and none of them were perfect :)

https://liftgammagain.com/forum/index.php?threads/lg-oleds-3d-lut-profiling-using-lightspace-thread.13105/#post-131512
Yes, agreed, but again, they talk about SDR, we don't! :)

I guess this only applies to SDR
That's the point!

Edit: Isn't this also what Calman does? I'm not sure there is manual HDR/DV calibration, only measurements and straight into a proper 1D/3D LUT combo uploaded to the TV.

Yes, Calman does a unity 1D LUT and 3D LUT for SDR, HDR, and DV.
... But they set a unity 1D LUT and then calculate greyscale corrections in that 1D LUT. Then they run a matrix LUT (or lightning I don't know what it's called these days) to create a primary 3D LUT. That's if you use the autocal workflow, anyway.
Yes (I haven't tried DV):
- they do ddc reset at first
- they do Autocal for 1dlut (with unityt 1dlut in place): measuring 20 points with LG 2018 Oleds (0 1.4 2 2.7 4 5 7 10 15 20 25 30 35 40 45 50 60 70 80 90 100 %), they start with 100% and going backwards; estimating 1 point at a time, uploading the 1dlut, then measuring that 1 point, adjusting/uploading/measuring etc., then move on to the next. So it's essentially the "same" as what we do with our 1dlut, the big difference is that it doesn't iterate over the whole 20 points at all that we can do if necessary (plus you won't have an uploadable file in the end of the process, it's handled internally, so I'm not sure you can upload the same calculated 1dlut multiple times, I guess not, and that's a big bummer)
This takes about 20 minutes with madTPG
- then they measure (with the new uploaded 1dlut in place) 5 colors (RGBWB) for HDR10 and create a 3dlut out of those (but what settings they use for this 3dlut???) then upload it
The problem with this part is, again, who knows what color patches they use (100% in gamma space?) and the resulting gamut can (will?) be seriously reduced, resulting in completely different colors than our "reference" factory preset.
- note that all of these steps are optional in the workflow (so these are the suggested steps, but you can skip any of these)

They claim the beta version that's been released in the last week works around/fixes the raised black and other issues with uploading a 1D LUT.
Well, after experimenting a lot with 1dlut, there's no magic trick here, they can only modify the shape of the resulting gamma curve, nothing else.

SamuriHL

4th April 2021, 15:32

Yes, for SDR. But we are talking about HDR, that's why I said that it's far from simple and maybe no one knows how it's supposed to work.

This isn't so hard as you make it out to be. By default the TV ships with a few different color temps set at the factory. Most people recommend to use warm2 because it's closest to D65 out of the box. The thing is, you have to remember that for HDR, internally the panel is converting 2.2 gamma to PQ EOTF, so setting the white point to 100 NIT D65 (or an AWP if you have the proper equipment to do so) is correct because you're measuring it at 2.2 gamma when you calibrate it. If you use a higher NIT white point, I suspect that's going to seriously mess with the TV's internal tone mapping.

As far as I know it's not Light Illusion who provided those extra tools :) But do they offer 1dlut upload capability now? (They didn't a year ago.) That means you have to accept everything they say, you can't try it for yourself.
And there's a valid usecase for 1dluts: e.g. to fix black crush (caused by panel replacement/firmware/etc), again with HDR.

Huh? DeviceControl has always had the ability to upload a 1D and 3D LUT as far as I know. I've been using it for over a year now. We just don't use 1D LUT for SDR or HDR.

That's absolutely true :) I've created around 10 1dluts for now and none of them were perfect :)

Yes.

Yes (I haven't tried DV):
- they do ddc reset at first
- they do Autocal for 1dlut (with unityt 1dlut in place): measuring 20 points with LG 2018 Oleds (0 1.4 2 2.7 4 5 7 10 15 20 25 30 35 40 45 50 60 70 80 90 100 %), they start with 100% and going backwards; estimating 1 point at a time, uploading the 1dlut, then measuring that 1 point, adjusting/uploading/measuring etc., then move on to the next. So it's essentially the "same" as what we do with our 1dlut, the big difference is that it doesn't iterate over the whole 20 points at all that we can do if necessary (plus you won't have an uploadable file in the end of the process, it's handled internally, so I'm not sure you can upload the same calculated 1dlut multiple times, I guess not, and that's a big bummer)
This takes about 20 minutes with madTPG
- then they measure (with the new uploaded 1dlut in place) 5 colors (RGBWB) for HDR10 and create a 3dlut out of those (but what settings they use for this 3dlut???) then upload it
The problem with this part is, again, who knows what color patches they use (100% in gamma space?) and the resulting gamut can (will?) be seriously reduced, resulting in completely different colors than our "reference" factory preset.
- note that all of these steps are optional in the workflow (so these are the suggested steps, but you can skip any of these)

Yes, I played with this workflow for months for both HDR and DV and was never able to get satisfactory results from it. I can't test the beta so I can't comment if it's improved or not. And you can actually measure more than 20 points if you want during greyscale (I think it's like a 42 point option or something like that). This didn't seem to make much of a difference for me. I got the same screwed up results regardless. Leon even showed us how to do a manual greyscale calibration through the DDC and while that was better it was still flawed. And extremely time consuming to boot.

Well, after experimenting a lot with 1dlut, there's no magic trick here, they can only modify the shape of the resulting gamma curve, nothing else.

Then they'll not be truly fixing the issues caused by using a 1D LUT, and yes, this applies to HDR as well as SDR.

chros

4th April 2021, 17:30

The thing is, you have to remember that for HDR, internally the panel is converting 2.2 gamma to PQ EOTF, so setting the white point to 100 NIT D65 (or an AWP if you have the proper equipment to do so) is correct because you're measuring it at 2.2 gamma when you calibrate it. If you use a higher NIT white point, I suspect that's going to seriously mess with the TV's internal tone mapping.
Yes, but that's the question, do we actually know how LG (in a given year) deals with HDR? I don't think so ...
(E.g. Vincent stated that he calibrates HDR10 to D65@203 nits, but how/why/with which method ... ? :))

Huh? DeviceControl has always had the ability to upload a 1D and 3D LUT as far as I know. I've been using it for over a year now. We just don't use 1D LUT for SDR or HDR.
Oh, I'm sorry, you are absolutely right, I forgot about it completely :o (I haven't used it since aiopylgtv is available).
OK, so the possibility was/is there for the gurus as well to play around with it.

SamuriHL

4th April 2021, 17:36

Yes, but that's the question, do we actually know how LG (in a given year) deals with HDR? I don't think so ...
(E.g. Vincent stated that he calibrates HDR10 to D65@203 nits, but how/why/with which method ... ? :))

Vincent uses Calman. As for how he'd calibrate D65 to 203 nits, that I don't know. I would wager a guess that he's only doing that on newer LG panels, not our 18 models, though. Why? Because with the newer models you can upload a custom tone curve. I'm not sure how to do this in the Calman workflow. Besides, he's got a pro license and isn't using Calman home.

Oh, I'm sorry, you are absolutely right, I forgot about it completely :o (I haven't used it since aiopylgtv is available).
OK, so the possibility was/is there for the gurus as well to play around with it.

Yes, you can definitely create and upload a 1D lut should you choose. And because of how LightSpace/ColourSpace work, you can actually compare having it uploaded with simply applying it from LS/CS and running measurements. And they found there's most definitely a difference between having it uploaded vs simply applying it like an external LUT box would.

quietvoid

4th April 2021, 19:10

I just got my meter and gonna try to do a simple SM whitebalance calibration.

I tried doing a ddc_reset without uploading the unity 1D LUT and my red saturation is way lower. Is this normal?
Just aiopylgtvcommand <IP> ddc_reset hdr_cinema False

I was expecting it to be the same but apparently not..

SamuriHL

4th April 2021, 20:21

I actually don't know what a DDC reset without resetting the 1D LUT to unity would do. I don't imagine that would be correct, though. I don't know enough about the software you're using to do that.

quietvoid

4th April 2021, 21:38

Yea, I just ended up doing a factory reset again.
I wasn't too sure how to calibrate the WP through service menu so went ahead and did 20 points SDR :D

Considering it was my first attempt I think I did okay. https://postimg.cc/YhG7mHXG
The TV by default was pushing red a lot..

But then I left Windows native and did a measurement in my usual VM, and measurements are much faster but also way off.
Oh I'm silly, the colorimeter is not even passthrough by USB..

For the SM, am I supposed to measure with HCFR/DisplayCal while the service menu is showing the white screen?

SamuriHL

4th April 2021, 21:58

No, you should be using HCFR/DisplayCal to show a white patch and turn off the IRE or whatever the hell it's called in the SM.

https://www.avsforum.com/threads/lg-oleds-3d-lut-profiling-using-lightspace-thread.3043520/page-74#post-60048596

This shows the process. You should be able to do it with HCFR/DisplayCal the same way.

EDIT:

Meant to add this. The reason these panels always push red is that is their native color. There's a reason we leave red at 192 in the SM and only adjust blue and green. We lose luminance otherwise.

quietvoid

4th April 2021, 22:40

Ahh I missed the part about the pattern :)
So far all my measurements were done in native gamut, I think that's what is recommended. Though I've switched to Auto afterwards.

Also all colors are way off with native, so I'm still not sure if greyscale should have been done in native or not for SDR.

SamuriHL

4th April 2021, 23:22

Once you upload a 1D lut or reset it I think it locks the gamut but I'm not entirely sure. Been a long time since I messed with the 1D lut. Then again you did say you didn't touch the lut itself only DDC reset but I'm not 100% sure if that doesn't automagically put it in unity. I truly don't know.

quietvoid

4th April 2021, 23:32

I did a factory reset after the ddc_reset mistake, so this is all starting from factory.

SamuriHL

4th April 2021, 23:48

I'm not sure. Once you do the service menu white point balance adjustment, in gamma 2.2 to 109% white, you should then be able to set the temperature you calibrated in your picture mode. At that point you should be able to calibrate that picture mode however you'd like. For HDR the gamut should be set to auto. Uploading a LUT will override that and grey out the gamut setting because it's handled by the LUT you create.

quietvoid

5th April 2021, 01:29

So far I've reset my 20 point stuff back to default.
Did white balance in the service menu, then back to 20 points which was simpler because less to do. Then grayscale luminance adjustments.

I didn't touch LUTs yet but I think I might need to for correcting colors, as adjusting primaries/secondaries seems to make things even worse.

Tried playing with madTPG but HDR isn't staying enabled. Shrug.

SamuriHL

5th April 2021, 02:53

madTPG can be temperamental at times. It's one of the reasons I switched to using pgenerator on the rpi. And why I got the HD Fury.

chros

5th April 2021, 10:57

I just got my meter
Nice! :) Don't forget to select the meter WOLED meter correction spectral file for it (in HCFR and in DisplayCal as well), WOLED 2016 is also good, but here is one for 2018 models (https://we.tl/t-rH9EnkLLxw) (up for a week), created by janos666 with i1Pro2 in SDR calibration mode after ddc_reset.

I tried doing a ddc_reset without uploading the unity 1D LUT and my red saturation is way lower. Is this normal?
It just then removes the 3dlut in SDR (ddc_reset (https://github.com/bendavid/aiopylgtv/blob/master/aiopylgtv/webos_client.py#L1321)), but I haven't recognised much difference with/without 3dlut in sdr, although I haven't measured it.

So far all my measurements were done in native gamut, I think that's what is recommended. Though I've switched to Auto afterwards.

Also all colors are way off with native, so I'm still not sure if greyscale should have been done in native or not for SDR.
So far I've reset my 20 point stuff back to default.
Did white balance in the service menu, then back to 20 points which was simpler because less to do. Then grayscale luminance adjustments.

I didn't touch LUTs yet but I think I might need to for correcting colors, as adjusting primaries/secondaries seems to make things even worse.
About SDR for video:
- as SamuriHL mentioned, the easiest way is to remove only the 3dlut, then create a 3dlut for madvr or the TV with DisplayCal (using madvr SDR preset, ~1 hour) (personally I only use madvr for SDR, nothing else)
- - you only have to deal with the 1dlut if you have serious issues with it for whatever reason (the big 3dlut can handle the grayscale as well)
- - for this, put the preset into Wide gamut (native)
- - to be able to upload into TV, you have to set Iridas format, full range (input and output), with 33 points for C8
- - you can do verification when it's done (using XXL video charts, 755 patches, that includes ~25 point for grayscale as well) and with HCFR

The good thing about this approach is that we will get 100% rec709 coverage (since we have ~96-97% dci-p3), so there's no gamut distortion. (That's why I don't see the point of uploading a dci-p3 one into bt2020, because at that time we will have it.)
Obviously the desktop and other application will be off with this if you only created a 3dlut for madvr, but I don't care about those.

And just to make it more confusing, the display produce different colors: PC vs non-PC mode, limit vs full range :)

Tried playing with madTPG but HDR isn't staying enabled. Shrug.
Yes, I get the same, I think it's because of driver + OS combination, it worked flawlessly 1 year go here :)
I mentioned the solution here (https://forum.doom9.org/showthread.php?p=1939693#post1939693):
I used madTPG as TPG, but even setting this up is problematic but only when using HDR10 output!:
- using the old driver (see my signature) that keeps the display in HDR even when app is not fullscreen
- run a calibration pattern video with metadata 1000/400 (maxCLL/maxFALL) and P3 primaries that triggers the display into HDR10 mode and stays in it (you can minimize the player to the tray)
- you can set a black image as background on the TV
- run madTPG for SDR calibration! (so HDR checkbox is off in HCFR, same in other software, e.g. Calman)

chros

5th April 2021, 11:15

Just one more thing about the HDR10 results (non-PC mode + limited (Low) signal):
- cinema HDR preset (https://imgbox.com/g/D1etg6dU4M) with factory luts and after full ddc_reset (where RGB balance is way off)
- cinema HDR preset after Calman (https://imgbox.com/g/HMa9xyBwNB) 20 points 1dlut autocal + matrix 3dlut
- - it also introduced other visible color errors on images
- - and I was wrong, you can actually see/modify what values Calman set for 1dlut and 3dlut (https://imgbox.com/g/e9fbF5qPK6) as well (the latter is huge, so not sure how useable this is)

So, maybe I should try to create a 40 points 1dlut for HDR10, maybe ... :)

Edit: and there's 1 more interesting thing about the gamma for HDR10:
- we see how it looks like with factory cinema preset (rotated S cure?)
- when I created a flat gamma 2.2:
- - it removed the fog/veil from dark scenes (lower end of gamma, below 30%)
- - but it introduced the same (fog/veil) in normal scenes!!! (other part of gamma, above 30%?)
- - so maybe I should aim, e.g. gamma 2.22 below 30% and gamma 2.1 above 30%?! :)

SamuriHL

5th April 2021, 13:09

I had the same thought about a 40 point calibration of greyscale when I was using calman. Thinking that measuring more points would make the lut more accurate. In fact quite the opposite happened. It created more artifacts, not less. It's entirely possible because of the limitation of the meter in low light conditions that it introduced more errors but I've no real idea. I know a user in the colourspace forum sells a patch set that was optimized for the LG oleds and he focused it more in the middle. Meaning less readings at the low end so that colourspace would calculate the lut values from the more accurate readings in the middle. I've seen people say that worked really well. Almost like a synthetic lut in a way. You can't really tell calman to do that though. I suspect you could with hcfr or displaycal?

Sent from my SM-G998U1 using Tapatalk

quietvoid

5th April 2021, 15:57

That's a lot of info @chros.
So I should only be clearing the factory 3D LUT if I was to replace it, which I haven't done yet.
For me, removing it reduces saturation a lot but I haven't measured if it's more accurate that way.

I guess I'll have to redo everything without 3D LUT, and upload a proper one after calibration.
Otherwise my colors are still not accurate.

What about calibration mode, should I be using it while measuring in HCFR/DisplayCal?
I also don't really know if I can just use a 3D LUT (and upload to the TV) instead of doing 20 points, since you mention it also corrects the greyscale?

SamuriHL

5th April 2021, 16:19

It all depends on if you're going to do a 1D lut or not as to what process you use. Here's how I do it:

-Calibrate White Point in the SM
-Reset the 3D LUT to unity, then disable calibration mode
-Measure and generate a 3D lut....this will include greyscale measurements
-Upload the 3D lut to the display

Calman measures while in calibration mode and I'm not necessarily sure that's the best of ideas. Calibration mode does change how the display is working. They do it because that's the only way they can mess with the DDC. But ideally you want to measure what you'll be watching.

quietvoid

5th April 2021, 17:01

Thanks, looks like I'll need to get comfortable with DisplayCal after all.

SamuriHL

5th April 2021, 17:38

If you are going to use a 1D lut then you can just add to the process above:

-Reset the 1D lut to unity, then disable calibration mode
-Measure and generate the 1D lut....upload the 1D lut to the display
-...repeat for 3D lut

I'd take it out of calibration mode after resetting to unity before measuring. Definitely do the 1D lut first, then the 3D lut as the 1D lut will correct greyscale requiring less correction (and potential for issues) in the 3D lut. In theory. :D

chros

5th April 2021, 17:52

It's entirely possible because of the limitation of the meter in low light conditions that it introduced more errors but I've no real idea.
This is with SDR again, right? :) It can happen with it, but with HDR10 even 5% is easily readable.

So I should only be clearing the factory 3D LUT if I was to replace it, which I haven't done yet.
Yes, in both SDR and HDR10.

What about calibration mode, should I be using it while measuring in HCFR/DisplayCal?
In SDR to create only 3dlut follow the process that I and SamuriHL mentioned:
-Calibrate White Point in the SM
-Reset the 3D LUT to unity, then disable calibration mode
-Measure and generate a 3D lut....this will include greyscale measurements
-Upload the 3D lut to the display
This means measurements for 3dlut happens in non-calibration mode.

But for HDR10 (also DoVi) I did everything (1dlut measurements) in calibration mode. As you can see in calibration mode with our 2018 sets the whole image gets brighter both in SDR and HDR10, but not in DoVi :) (I was told that 2019 sets works like DoVi in all modes, brightness won't change.)

I also don't really know if I can just use a 3D LUT (and upload to the TV) instead of doing 20 points, since you mention it also corrects the greyscale?
For SDR yes! :) But not for HDR10 :)

Thanks, looks like I'll need to get comfortable with DisplayCal after all.
That's easy for SDR, thanks to Florian, just select madvr sdr profile, select white drift on first page, then on 3dlut page select select pure gamma 2.2 or 2.4 (depending on what you aim for), wait and done, you are ready to verify it, e.g. looking at different scenes with it :) (You don't necessarily need the full calibration part (https://www.avsforum.com/threads/madvr-argyllcms.1471169/post-60182209).)

SamuriHL

5th April 2021, 18:11

This is with SDR again, right? :) It can happen with it, but with HDR10 even 5% is easily readable.

I really don't know what this means given that black is black. Low light means exactly that. Are you suggesting HDR has no low light measurements? I'm talking WELL below 5% when I say low light. We're talking 2% and lower. This meter does not handle that well.

But for HDR10 (also DoVi) I did everything (1dlut measurements) in calibration mode. As you can see in calibration mode with our 2018 sets the whole image gets brighter both in SDR and HDR10, but not in DoVi :) (I was told that 2019 sets works like DoVi in all modes, brightness won't change.)

Yea, it's not SUPPOSED to do that but it does. Which is why I measure in non-calibration mode cause otherwise you're measuring something that's artificially bright.

chros

5th April 2021, 18:30

Are you suggesting HDR has no low light measurements? I'm talking WELL below 5% when I say low light. We're talking 2% and lower. This meter does not handle that well.
I'm suggesting that 5% in SDR non-calibration mode is darker than the same in HDR10 calibration mode.
Yes, 2% and under it problematic in both.

Which is why I measure in non-calibration mode cause otherwise you're measuring something that's artificially bright.
Again, in SDR.

@quietvoid, 1 more thing about SDR, I use:
- ~120 nits (33 oled light) with gamma 2.4 for ISF dark mode, not 100 nits
- ~140 nits (39 oled light) with gamma 2.2 for ISF bright mode (just below the ABL level)

aron7awol

6th April 2021, 15:12

Apologies for such an on-topic post (I'm totally kidding and love to see anything related to technical aspects of OLED in here)...

I've been away for much of the custom curves testing, although I've been trying to at least keep up to date reading the relevant threads. Kudos to @quietvoid for his hard work doing a ton of testing specific to us in that area.

I've seen mentioned in the main thread that 128 is considered more of a daily driver than 130, although I admit I haven't done enough research to discover why exactly. Considering on most 1,000 nit content there is very little TMing necessary, and that when I sit down to actually watch a movie I want a consistent non-distracting experience above all else, I have been hesitant to upgrade because of mentions of brightness pumping.

So is there a combination of version/settings that will give the great results that @quietvoid has worked so hard to dial in for us, while avoiding something distracting like brightness pumping?

quietvoid

6th April 2021, 15:38

130 is a build with curve normalization, to avoid conflicting settings while testing.
It supposedly affects brightness a lot, which is why some are still on 128 for watching.

As for brightness pumping, madshi said there was no protection currently so I don't expect any of the latest builds to be perfect.
It's probably worse for lower nits than for us, I don't really notice much and I doubt it would happen in regular content.
If you want consistency I'd suggest avoiding the new curves completely.

Last I posted my settings: https://postimg.cc/vDKX0L21
Since then I think I changed my DTN down to 65, to be closer to what I saw in Dolby Vision.
And still using my R18 curve (signature).

chros

6th April 2021, 17:31

Indeed you did a good job and keeping the spirit alive, that's why I try to help you with anything I can :)

SamuriHL

6th April 2021, 18:54

Seriously I've been using 130 and your latest curve for most of my daily driving watching. Yes, there are brightness issues but as mentioned for us they're not quite as bad or noticeable. You did a really great job chasing that configuration down.

quietvoid

7th April 2021, 06:11

:)
I still hope there can be improvements in the curves, not too sure what madshi has planned though.

I ended up doing a quick DisplayCal calibration for SDR (175 patches, 3D LUT only), but the final results were a bit lacking for grayscale.
Did a 20 point on top but it was much more complicated, so I don't know if I should have.
Still improved on the verification by a lot.

Just gotta look at actual content and see if there are issues.,
So far I just noticed some darker shadows, but it's hard to tell which is more accurate, since it's a trade-off between detail and near-black artifacts.

SamuriHL

7th April 2021, 12:18

175 patches is very small. When I do sdr calibration I use just under 5000 patches.

Sent from my SM-G998U1 using Tapatalk

chros

7th April 2021, 12:31

Yes, do at least the default number (~1500) or higher.

quietvoid

7th April 2021, 13:50

The default was auto optimized and 175. Guess I'll have to redo it..
Not sure how it would improve the greyscale though, everything else was below 1.5 dE already.

SamuriHL

7th April 2021, 14:08

It all depends on what patches were included in the 175. It's likely greyscale will improve more with more patches, however.

chros

7th April 2021, 14:44

The default is 1500 patches for sdr, did you select the right profile, madvr sdr?

quietvoid

7th April 2021, 14:45

No, because it doesn't do an IRIDAS LUT by default and was only 16-235.
I want the calibration for all inputs.

chros

7th April 2021, 23:02

Okay, then select madvr sdr profile, leave everything on deafult, only select Iridas full range on 3dlut tab along with gamma 2.4. That's it.

quietvoid

9th April 2021, 05:24

Got bored and decided to bump up R18's brightness.
Very small difference for 2230/10000 curves at 870 DPL. Most noticeable is probably 10 000 nits content as usual.

chros

9th April 2021, 15:17

If we only had a usable build... :) Instead, we have:
- mexican soap opera with lot of drama, no action :D
- or a docu about early 1900 when the angry mob wants to lynch the factory owners :D

chros

9th April 2021, 16:33

I've measured DoVi PQ (https://forum.makemkv.com/forum/viewtopic.php?p=107747#p107747) if someone is interested.

VBB

9th April 2021, 18:42

What a mess over at AVS, but that's not surprising. I can appreciate the entire tone-mapping effort, of course, but the vast majority of people over there have only been around for the past two years or so. They have absolutely no clue about the years and years of algorithm testing here at doom9, and all the guys who were involved in that. And those are not the ones complaining right now. As much as I like AVS, I would have much preferred if all things madVR had stayed right here.

...and whoever at AVS made a certain user a super moderator, I have no words :P

chros

9th April 2021, 22:32

You are right. And who are you referring to?

VBB

9th April 2021, 22:47

Oh, I'd rather not name names, but he's been posting in that "Future of madVR" thread, and quite often in the HTPC section. huhn most certainly knows who I'm talking about ;)

SamuriHL

10th April 2021, 00:22

Yea, both those threads are a mess. Some sane people are simply asking for what the future may hold with the expiring beta builds in terms of being able to have a semi-updated, stable build of madvr. E.G. build 130 not super stable for most people's daily driving, but, 128 is going to expire. What do you do with that? And that's really all that was originally being asked. Now that's turned into a total CF for no real reason. SIGH

VBB

10th April 2021, 01:11

I'm not blaming you for chiming in over there, and it did finally prompt a great response from Mathias :)

SamuriHL

10th April 2021, 01:57

I just wanted to try to bring some sanity to it cause it was quickly going off the rails. And yes, I am happy with the response we got. I hope after his current Envy endeavor he'll make this a priority even above further tone mapping tweaking because it will set us up for being able to choose what builds we wanna use.

quietvoid

12th April 2021, 22:15

I ran a 1.5k patches calibration. It didn't improve the greyscale and introduced some 25+dE colors.
So my initial medium/175 patches along with 20 points still gets better scores on the XXL verification set.

Not sure what went wrong but I don't really see a point in redoing it since the first one is good.

chros

15th April 2021, 19:32

Something is/was wrong then, either during measurement or verification. You can upload the settings as screenshots and the verification html file. It should be below 4 dE.

chros

25th April 2021, 11:23

Guys found the meaning (https://www.avsforum.com/threads/oled-tvs-technology-advancements-thread.681125/post-60685673) of TPC, GSR, etc.
CPC is interesting: is that the standard ABL? Do we see its effect with Cinema DV preset with the glowing? But it can't be a power hog at that low luminance level...

Tronic

26th April 2021, 18:10

I have HDR working great with HDR10 BT.2020 limited range mode on LG C9, Windows in HDR mode, Nvidia drivers set to full range, CUVID decoding and MadVR display settings set to limited range, calibrated BT.2020 with BT.709 curve, HDR pass-through.

However, whenever playing any SDR videos (8 or 10 bit), black becomes gray even though MadVR's source and display are both set to TV levels.

OSD with HDR10 source:

OS HDR, 12 bit, RGB, full
D3D11 fullscreen windowed (10 bit)
P010, 10 bit, 4:2:0
HDR 992 nits, BT.2020
limited range (says upstream)

OSD with SDR 8 bit source:

OS HDR, 12 bit, RGB, full
D3D11 fullscreen windowed (10 bit)
h264, 8 bit, 4:2:0 -> NV12, 8 bit, 4:2:0
limited range (says upstream)

Problem partially solved:
MadVR presents the SDR source as 10 bit SDR output, after which Windows does its HD Color mapping: this takes the colors in PC range, and that's why it turns gray.

I think I'd prefer for MadVR to convert it into HDR10 format, identical to that used in true HDR movies, and then avoid Windows doing any color conversion.

quietvoid

26th April 2021, 22:11

I think you just need to not use Windows' HDR?
And leave madVR to enable HDR output (through the GPU APIs directly), with profiles if needed.

madVR does not convert SDR back to HDR10.

Tronic

27th April 2021, 20:06

I prefer to use HDR mode all the time because that allows Youtube HDR and other such things to work on desktop, and also because Windows color mapping in that mode is very good and there's a nice slider in HD Color options for desktop brightness.

But I've now setup two profiles, which work great:

"HDR": TV, BT.2020
"SDR": PC, BT.709 - but still report BT.2020 to display!

No gamma processing because there is no support for sRGB in MadVR and because the untouched gamma already looks good (might even be precisely correct, I cannot tell).

Automatic switching code:

if (HDR) "HDR"
else "SDR"

Asmodian

27th April 2021, 20:44

You run your desktop in HDR mode all the time?

That is a lack of worry about burn in that I can respect (but not emulate)! :scared:

SamuriHL

27th April 2021, 22:30

Seriously lol

Tronic

28th April 2021, 06:37

You run your desktop in HDR mode all the time?

That is a lack of worry about burn in that I can respect (but not emulate)! :scared:

In Windows HD Color settings there is a slider to control how bright SDR content appears when in HDR mode, so it doesn't burn your eyes (or the screen in case you worry of that). Dark mode UIs on desktop and websites also help!

Display burn in I am absolutely not scared of. TVs evolve at such a rate that I wouldn't keep one for more than a few years anyway. A few years back you couldn't get 4K/120 or VRR, and soon enough I'll be replacing the current C9 with a bigger screen as the insanely high prices of those come down. Maybe the peak brightness keeps improving, and 8K panels certainly are coming in (in case anyone can find use for that resolution). Good times :)

Asmodian

28th April 2021, 07:51

I use a decent SDR calibration, I cannot handle the Windows SDR->HDR look. :(

and 8K panels certainly are coming in (in case anyone can find use for that resolution). Good times :)

I can see a really good use for a 75"-85" 8K screen as a desktop monitor. That would be a really nice working surface. It won't be that long. :)

Tronic

28th April 2021, 14:55

I use a decent SDR calibration, I cannot handle the Windows SDR->HDR look. :(

Out of curiosity, which gamma curve do you calibrate the desktop for? I find all desktop stuff using different curves anyway.

For a long time most games just ignored gamma or hardcoded 2.0 (x*x and √x). New games have configurable gamma values (defaulting to 2.2) and some might even be using the sRGB gamma ramp (linear region for the darkest shades). (And luckily a select few now support proper HDR.) Then, I am pretty sure that most video material is actually BT.709 and no-one bothered the difference between that and sRGB. Photographs... Maybe sRGB if processed with good software, or in rare cases might even include colour profiles for extended gamuts. Website and UI colors: whatever happened to look good on the uncalibrated display of the guy who designed it (not to mention that Mac and PC always were quite different).

I still couldn't find out which gamma ramp Windows SDR-to-HDR is using, but judging from the slight visual difference to the sRGB which I am used to, I presume BT.709 (should buy a calibrator and measure). At least the primaries seem to be BT.709, which fortunately are identical to those of sRGB. I have forced standards mode in Windows (i.e. use BT.2020 on TV all the time).

SamuriHL

28th April 2021, 15:40

I have a picture mode for 2.2 and another for 2.4 depending on what I'm viewing and when.

Asmodian

29th April 2021, 01:36

Then, I am pretty sure that most video material is actually BT.709 and no-one bothered the difference between that and sRGB.

What would you want to do differently between BT.709 and sRGB content? There is nothing different between them that should be bothered about.

You never calibrate your display to the BT.709 or sRGB gamma curves. Always use BT.1886 or pure power, even theoretically, for calibrating a display for BT.709 or sRGB content. The near black linear region in those gamma curves is intended to deal with issues converting the early digital video data to display on a CRT. They are basically hacks to do something like BT.1886 without any specifics of the display and should never be used as calibration targets.

That said, whatever transformation Windows does to compress the SDR desktop into the the bottom of the HDR signal is very far removed from a proper calibrated 2.2/2.4/BT.1886 SDR image on my C7/C9/CX. It isn't even trying to get close as far as I can tell.

I have a picture mode for 2.2 and another for 2.4 depending on what I'm viewing and when.

Yeah, me too. I usually use 2.4 (dark room).

SamuriHL

29th April 2021, 04:16

Daytime viewing for me requires 2.2. Movie nighttime viewing is done at 2.4. That works out well for me.

quietvoid

9th May 2021, 04:09

Spent the whole day fixing my HDR. Looks way better now.
@chros if you don't mind: https://hub.displaycal.net/forums/topic/lg-c8-lut/page/18/#post-30095

As for madVR, I had a quick look at R19 vs single steepness 600 and to be honest I'd be fine with either, there really is not much difference.
Guess we'll know more with the next build.

SamuriHL

9th May 2021, 12:47

Assuming the next build changes anything. It could just be a new expiration. We can't assume anything given how busy he's been.

Sent from my SM-G998U1 using Tapatalk

quietvoid

15th May 2021, 21:01

I've been working on some formulas to make 2 lines based curves adaptive by DPL, and ended up making a couple presets. From very high to very low.
However my preference is still a curve very similar to R19, which is the "high" preset.

There is little difference in midtones brightness between presets at 870 DPL, and the highlight detail is pretty much the same most of the time.
Very low might leave more place for stronger contrast recovery but I didn't really test yet.

I'd post on AVS but madshi wants some final proposals and we're waiting for Fer15 to be able to test, because I don't know yet if the presets I've made are good enough for him.

I'm still using single steepness for actual watching for now (because of the artifacts I've brought up).
Otherwise my settings for 132: https://postimg.cc/dhKcm9gx

If you want to test the new bright & very low curves:
High: https://gist.github.com/quietvoid/cf1454ad05efba1ad34cf6963b55ca1a
Very low: https://gist.github.com/quietvoid/467e1bb8a08dcf58491e5f995fe43c8e

SamuriHL

15th May 2021, 21:26

Awesome. I'll try to get some tests in with these this week some time. I got Raya coming so that'll be a good one to play with. I've also been using single steepness 600 for movie watching lately. It'll be interesting to see how this latest information plays out.

Magik Mark

15th May 2021, 23:12

Thanks QuietVoid for sharing!

SamuriHL

16th May 2021, 02:45

I tried your new setting with your high curve. I like it. That looks very nice and well balanced.

quietvoid

16th May 2021, 14:28

It's pretty much the same as R19 was for 870 nits.

SamuriHL

16th May 2021, 15:19

I never got around to r19 lol

Sent from my SM-G998U1 using Tapatalk

chros

22nd May 2021, 02:39

Guys, have you seen this sh!tshow about the new EVO panel/ecosystem crap? :D
link (https://www.avsforum.com/threads/2021-lg-oled-c1-g1-owners-thread-faq-no-price-talk.3190432/post-60754597), link (https://www.avsforum.com/threads/2021-lg-oled-c1-g1-owners-thread-faq-no-price-talk.3190432/post-60754864)

VBB

22nd May 2021, 03:24

@ttnuagmada also has an account here, btw. Good job, Adam! We kind of expected the C-Series to be gimped by software this year, and this proves it.

SamuriHL

22nd May 2021, 03:51

Uhhhhh, what the??? That doesn't seem good!

VBB

22nd May 2021, 04:11

Well, the fact that the C1 can be un-gimped is great, but why LG chose to do this to begin with is beyond me. They should have kept the same picture quality across the lines, just like all the other years. Too late now, but the real sh1tshow, IMO, is the fact that you can buy a C1 right now and not know what type of panel you'll end up with (see the C1/G1 thread over at AVS for more info). That brings us full circle to the original meaning of the term 'panel lottery', namely when Samsung started using different manufacturers' panels into its same model/series TVs. And that is total BS, if you ask me.

SamuriHL

22nd May 2021, 04:51

It really is and gives me pause about buying anything new. Not that I was really in the market anyway but at some point I'm going to replace the plasma in the bedroom with my c8 and get a larger screen for the living room. Guess it's not this year.

Sent from my SM-G998U1 using Tapatalk

chros

22nd May 2021, 12:33

the real sh1tshow, IMO, is the fact that you can buy a C1 right now and not know what type of panel you'll end up with ... And that is total BS, if you ask me.
Yes, that's what I meant :D
I mean they could put the old panels into the crippled Bx/Ax series and not handicapping the precious Cx series ... And putting the new Evo panel into CXs as well?!!! :D ... what a mess ... it's just beyond me ...

Asmodian

22nd May 2021, 22:27

LG is in the process of converting all (most?) of their OLED lines into EVO lines. As capacity is transitioning the switch gets messy; they run out of older panels so they have to use the EVOs instead.

Since stockpiling last year's panels was never a realistic option, and they did not have enough EVO panels when the C1 was released, this seems like a reasonable situation. No one gets a worse panel than they were expecting, do they? I do hate panel lottery within one model, but at least the unexpected panel is better than the at release panel. :p

SamuriHL

22nd May 2021, 23:11

No thanks. I'll continue waiting. :D

Asmodian

22nd May 2021, 23:43

Yeah, waiting still seems wise unless you go for a G1. It would be sad to get a non-evo panel today. I want my deuterated materials, please. :)

Just that I prefer LG moving all manufacturing to the EVO line as fast as possible, to the point that I am even willing to forgive them putting two different panels in the C1. I am not ready to replace my CX anyway, so maybe it is too easy to not care since I wasn't in the market anyway. :p

SamuriHL

23rd May 2021, 00:27

Exactly. I'm looking at some day moving up to a bigger panel for the living room, but until LG gets their collective s*** together, I'll live with my C8.

Yes, I get that we all want them to move to EVO but this move is confusing as hell and not doing them any favors. They should have just moved everything to EVO from day 1.

Asmodian

23rd May 2021, 00:56

Well, yes, if reality would have allowed it for at anything like a reasonable price and/or in a reasonable timeframe... but that was never an option.

LG sells a lot of panels, these panel manufacturing plants are not small operations. It takes significant time and money to covert them to produce EVO panels, even as small as that change is. Maybe I have been working in a manufacturing related industry too long. :p

Intel should have released 10nm a long time ago too. :devil:

Is it really confusing? Some C1s come with the new panel and some do not. Is it more complicated than that?

SamuriHL

23rd May 2021, 01:15

Here's the thing about that theory.....LG makes ALL the OLED panels sold (at least for now...that's changing soon). Meaning panels for Sony, Panasonic, etc. Give them the old crap! And save the EVO panels for the LG lineup. Gives them a leg up, burns old inventory, and everyone (except their competitors) wins. :)

nevcairiel

23rd May 2021, 07:39

Contracts mean that their direct customers get exactly what they asked for. :p
Besides, you really can't complain for getting a TV in the specification it was meant to be in. If you want a guaranteed better one, need to wait for the next gen. Never know how long one was sitting in some vendors warehouse, independent of what LG does or doesn't do.

SamuriHL

23rd May 2021, 13:51

Which is why I'm waiting yes. [emoji16]

Sent from my SM-G998U1 using Tapatalk

ashlar42

28th May 2021, 21:53

LG choices this year are making me look the Sony A80J with real interest.

Asmodian

28th May 2021, 22:05

The A80J does look like a very nice display. If Sony releases a 48" I think I would buy it.

chros

29th May 2021, 08:13

LG choices this year are making me look the Sony A80J with real interest.
There are reports on Avsforum that the A80J also has mixed panels. (by measuring SPD)

Magik Mark

31st May 2021, 01:50

Hi chros!

Just curious. Why is it that your madvr setting hasn't been updated?

chros

31st May 2021, 07:19

Because I use b113 and I even disabled DTN :)

Magik Mark

31st May 2021, 07:59

Oh I see. Do you have anything against on later versions?

Magik Mark

1st June 2021, 11:26

chros,

I'm wondering if you have started playing around with CRU? I think the last setting doesn't apply anymore as nvidia change the timing for 23.976hz

chros

5th June 2021, 10:52

I created a new 1dlut for DoVi Cinema preset (https://hub.displaycal.net/forums/reply/30492/), the result is better than factory (although not prefect, yet?).
As a result, I started to watch 1080p SDR content with Oppo + DoVi again ... :)

@Magik Mark: I use those old values I posted back then, I don't upgrade anything until it's absolutely necessary :) (see my signature for details)

chros

5th June 2021, 12:27

Oh, and as it turned out, the new Sony oleds also suffer from elevated blacks and the same near black glowing in DoVi! :) Ridiculous... Calibrators advise is: reduce brightness or contrast... :D

heartog

8th June 2021, 09:40

Just got the G1 few days ago and noticed that marVR would switch my TV picture and sound modes when going into fullscreen windowed, does not happen with my old samsung NU8000 or when i use EVRCP. Using Potplayer

huhn

8th June 2021, 15:23

disable HDMI content type in nvidia control panel.

if you like you could play around with it to show how it changes the picture mode to like movie if selected.

con remember any other user than me with a Panasonic reporting picture mode changes.
sound modes is new to me.

chros

8th June 2021, 17:14

Yes, it doesn't do anything with 2018 LG models.

VBB

8th June 2021, 19:10

It didn't do anything before this year's LG series, but now it changes to Game Mode on the TV if the Nvidia control panel is set to AUTO or Desktop Programs. You'll have to switch it to Full-screen Videos to make it stick.

chros

9th June 2021, 15:32

Did you get a C1? :)
Interesting feature, not sure how useful it is: does Game mode in normal mode (non-PC) support chroma 4:4:4?
Vincent mentioned that it has banding in non-PC Game mode but it's fine in PC mode :D
I wonder whether they still have the cyan and near black posterization issues.

VBB

9th June 2021, 17:56

Nah, I'm still rocking my ancient C7, but I did help ttnuagmada (the guy who first tried the C1-to-G1 switch in the service menu) a while ago with this very issue.

chros

10th June 2021, 07:36

Nice! :)

Tronic

16th June 2021, 19:51

[LG C9] After upgrading to RTX 3090 I can no longer adjust smooth gradation (greyed out) or TruMotion (Vivid and Smooth modes are missing, and the User mode doesn't provide any smoothing, only OLED Motion). These settings are not available in any picture mode and many others are also greyed out. Any idea why this would be? I already tried disabling G-Sync (Nvidia drivers) and Instant Game Response (LG) but that makes no difference. Running the video player in 2160p23 HDR just as I did with the old GTX Titan GPU, but connected to HDMI1 with eARC, where the old setup ran through amplifier on HDMI2 (because eARC didn't work with the old GPU).

The missing settings are available on HDMI2 and Live TV.

Tronic

16th June 2021, 21:57

Had to rename the HDMI 1 input to something other than PC. Using the PC mode disables most settings that may still be useful for movies. Now with ALLM (Instant Game Response) enabled MadVR switches profiles so that desktop and games are in Game mode without filtering and movies get their smoothing.

Asmodian

16th June 2021, 22:03

PC mode is 444 chroma, but disables all the processing that is done in 422 chroma. I hate chroma subsampling but it is one or the other on these LG TVs. :(

ashlar42

18th June 2021, 13:45

Ended up getting a 77C1 (and a RTX 3070). It will be with me in a week or so. Prepare to be pestered with questions. :-)

Jokes aside, if there's any reading to be done, please point it out to me. I will try to read everything I find, but information is scattered all around the web...

chros

18th June 2021, 15:44

Follow tge links in my signature :)
Btw you can make a G1 out of your C1 if it has the new Evo panel :) (search on youtube how)

VBB

18th June 2021, 18:14

With the latest firmware, there is no difference any longer between the C and the G setting. LG caught on ;)

ashlar42

18th June 2021, 23:15

With the latest firmware, there is no difference any longer between the C and the G setting. LG caught on ;)
I bought C1 fully intending NOT to put it in G mode, firmware updates notwithstanding. I have read about the differences, coming from plasma I read about blinding brightness and I'm pretty sure that, when watching at night (as I usually do), I will need to lower OLED light setting.

VBB

19th June 2021, 01:42

Yeah, I would have done the same. I'm sure it'll be plenty bright for you. Enjoy your new toy!

SamuriHL

19th June 2021, 03:12

That's a nice rig. I'd like to move up to 77 with my next TV. I'm holding off a bit more probably next year or the year after. I did pre-order a vrroom just to be ready though.

ashlar42

22nd June 2021, 18:53

TV is fantastic. WOW!

Now on to problems, I hope you'll all be patient with me. I thank each and everyone of you in advance.

Sooo.. this might be a Kodi issue, a madVR issue, a Windows issue, I don't know.

If I start HDR content, the TV (an LG OLED C1) switched to HDR, playback is fine, all good. Problem is that when I stop the video, colors are all wrong in the interface (like oversaturated). If I close Kodi and restart, they go back being ok.

If I activate HDR in Windows 10, this doesn't happen, but everything is played back in a way that keeps the TV in HDR mode (even when in Kodi interface).

Edit: the other problem I have is I don't know where to find the "HDMI Input" type thing. I see people often talking about "renaming it to PC" and stuff like that but I can't find it in settings and I honestly don't grasp the implications.

chros

22nd June 2021, 20:34

No worries, just shoot the questions. If you will put around 100 hours into it, i'd like you to do some small tests for us, hope that's ok.

HDR issue is probably caused by "driver". Personally I can't help you with that since I use an old driver.
Either others can help or you have to experiment.

About PC mode: I don't know where it is in the new interface of webos, you had to push the input button on the old one for that. Search for C1/G1 videos on youtube, someone showed it for sure (I think Vincent as well).

chros

22nd June 2021, 20:35

With the latest firmware, there is no difference any longer between the C and the G setting. LG caught on ;)
Cheeky bastards :D

SamuriHL

22nd June 2021, 20:47

HDR is a pain in the ass. That issue is "normal" depending on the software in your video chain. Letting madvr switch HDR on and off is the best way, but, as you noticed, until madvr exits completely it doesn't necessarily reset. There's an option in madvr to reset the video after playback although I don't remember exactly what it's called off the top of my head. Check that it's set. Kodi probably has something like that built in, too. Make sure to pick one of them, not both. Personally I let madvr do it.

ashlar42

22nd June 2021, 21:38

No worries, just shoot the questions. If you will put around 100 hours into it, i'd like you to do some small tests for us, hope that's ok.If it's within my capabilities, I'll try. Nothing too complicated, I hope, as the HTPC is used by my wife and daughter too, so I have to minimize the risks of seriously messing it up.
HDR issue is probably caused by "driver". Personally I can't help you with that since I use an old driver.
Either others can help or you have to experiment.
I checked the drivers guide in this forum and there did not seem to be outstanding problems in new drivers.

About PC mode: I don't know where it is in the new interface of webos, you had to push the input button on the old one for that. Search for C1/G1 videos on youtube, someone showed it for sure (I think Vincent as well).
Ok, the other part of the question is: why is PC mode desirable?
HDR is a pain in the ass. That issue is "normal" depending on the software in your video chain. Letting madvr switch HDR on and off is the best way, but, as you noticed, until madvr exits completely it doesn't necessarily reset. There's an option in madvr to reset the video after playback although I don't remember exactly what it's called off the top of my head. Check that it's set. Kodi probably has something like that built in, too. Make sure to pick one of them, not both. Personally I let madvr do it.
I am looking at the rendering options in madVR and I can't find anything regarding resetting after playback.

Unless you're talking about these options:

https://i.imgur.com/3nyMJkA.png

Edit: while researching... I suppose this might be it: https://forum.doom9.org/showpost.php?p=1927705&postcount=2097

chros

22nd June 2021, 22:44

Maybe, if they haven't fixed it yet (that's why i use v398.18).
PC mode: you get chroma 4:4:4 with it.
About tests: nothing complicated, we are just curious whether LG fixed the previous issues. Let us know when you are ready, no rush :) What other external playback devices do you have? (BD player, shield, firestick, etc)

ashlar42

22nd June 2021, 22:47

Holy hell https://forum.doom9.org/showpost.php?p=1929660&postcount=2181

Six and half months and the problems is still there? Gonna check the latest drivers that came out today. I'm speechless... and not in a good way.

chros

22nd June 2021, 23:02

ashlar42

22nd June 2021, 23:05

Actually it's a year old at least, I noticed this starting with 442.92.
But it's not clear to me whether it's a driver issue (due to Windows 10 HDR API) or madvr needs to be updated.
That's why I suggested to get an old pascal card.Yeah, the 3070 was a gift, not a choice. Is it possible to install older drivers with a 3070? Or they refuse to install, not recognizing the card?

chros

22nd June 2021, 23:08

You can research it, I never tried, but I guess it won't work.
If you don't play games, you can sell it, you will get good price for it nowadays :)

nevcairiel

23rd June 2021, 00:27

But it's not clear to me whether it's a driver issue (due to Windows 10 HDR API) or madvr needs to be updated.
That's why I suggested to get an old pascal card.

madVR is quite out of date as it comes to OS HDR integration at this point. Other video renderers in other players manage the transition into HDR and back out perfectly.

SamuriHL

23rd June 2021, 00:28

I am looking at the rendering options in madVR and I can't find anything regarding resetting after playback.

Unless you're talking about these options:

https://i.imgur.com/3nyMJkA.png

Edit: while researching... I suppose this might be it: https://forum.doom9.org/showpost.php?p=1927705&postcount=2097

Yea those are the options. I had to step away from my machine for a bit so couldn't look them up. But that's what I use with MC28 and it works well for me.

SamuriHL

23rd June 2021, 00:30

madVR is quite out of date as it comes to OS HDR integration at this point. Other video renderers in other players manage the transition into HDR and back out perfectly.

Yes. nVidia even had to put a work around in the driver to get it to work at ALL for madvr. Not ideal and not likely to get fixed any time soon.

ashlar42

23rd June 2021, 00:32

madVR is quite out of date as it comes to OS HDR integration at this point. Other video renderers in other players manage the transition into HDR and back out perfectly.
Yeah, I came back to note how regular Kodi 19.1 has no problem switching to HDR and switching back, all without exiting from the application.

Quite a sorry state of things. And, at this point, it's not even Nvidia's responsibility, I'm afraid.

oldpainlesskodi

23rd June 2021, 08:16

LG OLED BX, madVRhdrMeasure113 (0.92.17) and AMD no problems with dsplayer and HDR.

Re PC settings on the new OS, press the input button on the bottom of the remote and select Dashboard, then click the settings icon on the top right and select you input - don't highlight the name, but the actual HDMI pic, and then scroll down and select PC.

I don't bother with PC mode, as it disables some useful options (to me).

ashlar42

23rd June 2021, 11:44

LG OLED BX, madVRhdrMeasure113 (0.92.17) and AMD no problems with dsplayer and HDR.
On what driver version? Latest? Or even AMD is playing the whack-a-mole game with madVR users?
Re PC settings on the new OS, press the input button on the bottom of the remote and select Dashboard, then click the settings icon on the top right and select you input - don't highlight the name, but the actual HDMI pic, and then scroll down and select PC.

I don't bother with PC mode, as it disables some useful options (to me).
Thank you. I actually suspect that it activated PC mode for me already, as I can't access several options (Clarity ones, mainly). I will report back on this.

oldpainlesskodi

23rd June 2021, 11:58

As per my sig, 21.4.1 and I have zero issues with HDR etc.

ashlar42

23rd June 2021, 14:09

huhn

23rd June 2021, 14:42

my guess would be ALLM.

try the HDMI content type and change to something you prefer.

oldpainlesskodi

23rd June 2021, 16:10

Any idea as of why my HTPC makes the 77C1 gray out Noise Reduction, MPEG Noise Reduction, Smooth Gradation, Cinema Screen and TruMotion? This on all video modes, apparently (tried Standard and ISF ones)
PC uses an Nvidia RTX 3070, Windows 10 x64 1809, latest drivers (updated today), certified HDMI 2.1 cable. Tried setting it up as PC input and regular one, no change.

Just to check - is your GPU set at 60 or 120 hz?

My LG is also hdmi 2.1, so will do 120 hz, but I leave it set at 60 (as I use 23,24, 25hz etc), so just a guess but 120 hz might force PC mode (which would actually make sense)....just a thought.

ashlar42

23rd June 2021, 20:29

GPU is set at 60Hz. It was the same when set at 120Hz. As I wrote, I had it set up as both PC and another type (did not check what but I supposed it’s only PC type that could make a difference). I’ll try other types...

Edit: it was GSYNC.

oldpainlesskodi

24th June 2021, 07:14

Glad you got to the bottom of it.

ashlar42

24th June 2021, 13:01

Glad you got to the bottom of it.
Yeah, if I have to be honest, I’m absolutely speechless at how little forethought went into helping customers understand what the TV does in any given situation.

For instance, Cinema Screen is greyed out no matter what.
Also, on Nvidia’s control panel there is now a setting that you use to “tell” the screen what kind of content you are sending. If you select games, all clarity options are greyed out. If you select movies, they are accessible and Cinema Screen is greyed out (in active position, don’t know if that means it will be always on, will need to check further).

All in all, such opacity in settings.... bleah.

oldpainlesskodi

24th June 2021, 13:04

yeah, I hear you...took me some time to figure out what was going on lol.

ashlar42

24th June 2021, 15:39

madVR is quite out of date as it comes to OS HDR integration at this point. Other video renderers in other players manage the transition into HDR and back out perfectly.I know of Kodi. What are the other renderers/players? Under Windows 10?

We could ask madshi to fix this. It seems relevant enough, it’s not a minor quibble from a couple of users...

SamuriHL

24th June 2021, 16:05

You could ask madshi, but, the reality is he's focused on improving the Envy. The only changes to madvr are to improve HDR tone mapping. So it's not likely to get fixed any time soon.

nevcairiel

24th June 2021, 21:55

Its really a shame how madVR has been left hanging there. It wouldn't be so bad if it didn't happen in that time of transition for HDR, and its not in a place to catch up right now. That, and that its basically DX9 on the inside makes it shows its age.
Reliable HDR would be a big gain, maybe something needs to be done about that.

SamuriHL

24th June 2021, 23:37

maybe something needs to be done about that.

Yes.....something. :D That would be very nice indeed!

huhn

25th June 2021, 01:40

mpcVR is in a good state.
adding some proper scaler and 3D LUT support and is not far of madVR and also be a head of madVR in other places.

there HDR support is working just fine no getting out of fullscreen and HDR breaks like with AMD.

oldpainlesskodi

25th June 2021, 08:48

I've no issues with dsplayer+madvr+amd+hdr, what issues are you/were having?

chros

25th June 2021, 17:45

Its really a shame how madVR has been left hanging there.
This...

kostik

25th June 2021, 18:12

Madshi: "Haven't had a chance to implement the curves from quietvoid and Fer15 yet. However, let's start by fixing whatever needs fixing first. This new build should fix the DVD crash, and it adds an option to enable/disable brightness equalization:

http://madshi.net/madVRhdrMeasure133.zip
"

https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/page-658

quietvoid

26th June 2021, 19:40

beta 134 with brightness equalization switch requires changing DTN to preference again.
I had it at 60 previously but it's too bright for me without brightness equalization, so I'm now at DTN 70.

SamuriHL

26th June 2021, 20:56

I'm hoping we see a 135 with your curves built in. That would be ideal. I just took a quick look at 134 and it's looking really good. Will definitely be the build I use until we get something better.

aron7awol

26th June 2021, 21:10

beta 134 with brightness equalization switch requires changing DTN to preference again.
I had it at 60 previously but it's too bright for me without brightness equalization, so I'm now at DTN 70.
Do you feel like this is now perfectly stable or are there still brightness pumping issues from time to time? In any case, I'm interesting in trying the settings you are using as your daily driver. Would you share?

quietvoid

26th June 2021, 21:17

Magik Mark

26th June 2021, 22:44

Thanks Quietvoid for the new setting

SamuriHL

26th June 2021, 22:59

I've not seen any flashes with the curves. I'll be watching a movie soon with 134 and your curves so I'll keep an eye out for it.

kostik

27th June 2021, 18:57

I haven't really watched much so I can't say for sure.
I think in most titles you wouldn't have an issue because they're usually not over 1500 nits.

This is what I have for settings now: https://postimg.cc/ZC5W2qPf
I can't use custom curves because I get flashes (might be caused by having FastBlur enabled for performance).
But steepness 600 is very similar for everything under 10000 nits.

Those curves also fit LG OLED CX? it will help to change the DPL to my measured peak brightness?

Its been long since I used madVR for HDR movies, I just watch via streaming from my PC because I am not sure if my HDR calibration can get ruined by madVR settings even when just outputting HDR using 'passthrough' option in MadVR , let alone using curves and other new DTM madVR settings. What do you think?

SamuriHL

27th June 2021, 20:21

You can't "mess up your HDR calibration" using madvr. It doesn't change anything on the TV. Yes, you should let madvr know your measured peak brightness and let it tone map to that.

madshi was mentioning something about being able to calibrate newer LG's (I'm guessing the C1 but I don't know for sure as I'm still on my C8) to use a clipping tone mapping calibration. That would absolutely rock since below the clipping point it wouldn't touch anything so that you'd be able to use JUST madvr tone mapping. I don't know how this works or how you can calibrate it like that. I guess Stacey Spears was experimenting with it. For the C8 we chose 870 nits after a ton of experimentation to see what curve the TV chooses. For your CX I just don't know enough to be able to help you configure it best for madvr. Given that the tone mapping can be configured I would say you have more options than we do. Maybe someone with a CX can chime in here.

But in any case, do feel safe knowing you can experiment in madvr and not mess anything up.

chros

27th June 2021, 20:48

This is what I have for settings now: https://postimg.cc/ZC5W2qPf
I can't use custom curves because I get flashes (might be caused by having FastBlur enabled for performance).
But steepness 600 is very similar for everything under 10000 nits.
Cheers, so I have set it up for the first time to use Your curves :) I'll report back in couple of days how it behaves.
Here's my settings for HDR output with b134 (https://imgbox.com/ypLdESTJ#).

I don't see any flashes for now with your curves (FastBlur enabled here as well), maybe higher nits contents are affected?

Those curves also fit LG OLED CX? it will help to change the DPL to my measured peak brightness?
... let alone using curves and other new DTM madVR settings. What do you think?
Yes, can be helpful for high nits content (> 1500?).
With curves, set DPL to 870 before importing, then import the curves and apply, then set back your DPL to a lower value. (I use 800 with a B8 as you can see).

madshi was mentioning something about being able to calibrate newer LG's (I'm guessing the C1 but I don't know for sure as I'm still on my C8) to use a clipping tone mapping calibration. That would absolutely rock since below the clipping point it wouldn't touch anything so that you'd be able to use JUST madvr tone mapping. I don't know how this works
I'm not aware of any settings like this (or just forget it) beside of HGIG (I think this one only available in game mode), or did they add a hidden setting for it recently?

or how you can calibrate it like that.
Now that's the interesting part :) I didn't want to reply there to this, because it would have opened up a whole new wormhole :)
Anyways, with this 3dlut enabled hdr output, madvr opened up a bunch of new possibilities. We all know that the panels are really unstable in hdr mode, so no one even tried (at least not seriously) to create a 3dlut for HDR, they just measure couple of colors and generate the 3dlut.
Let alone that LG has 1dlut and 3dlut in them, so I don't even understand what Stacy wants to experiment with: you would have to remove the 3dlut in the TV (1dlut can remain in place) and try to create one for madvr hdr output.
(But with recent nvidia drivers I guess even madTPG is screwed ...)

kostik

27th June 2021, 21:28

You can't "mess up your HDR calibration" using madvr. It doesn't change anything on the TV. Yes, you should let madvr know your measured peak brightness and let it tone map to that.

madshi was mentioning something about being able to calibrate newer LG's (I'm guessing the C1 but I don't know for sure as I'm still on my C8) to use a clipping tone mapping calibration. That would absolutely rock since below the clipping point it wouldn't touch anything so that you'd be able to use JUST madvr tone mapping. I don't know how this works or how you can calibrate it like that. .

I think Stacy Spears wants to do an experiment with Roll-Off points set to 100% and see how it works with madVR TM ( possible with Device Control or Calman on C9 and newer models) :

https://imgur.com/9P1myeH.jpeg

SamuriHL

27th June 2021, 21:51

I'm not aware of any settings like this (or just forget it) beside of HGIG (I think this one only available in game mode), or did they add a hidden setting for it recently?

Apparently what kostik is saying. I don't know any details. Only that madshi mentioned it in the AVS thread when he released 133 and everyone discovered that it completely borked HDR output.

Now that's the interesting part :) I didn't want to reply there to this, because it would have opened up a whole new wormhole :)
Anyways, with this 3dlut enabled hdr output, madvr opened up a bunch of new possibilities. We all know that the panels are really unstable in hdr mode, so no one even tried (at least not seriously) to create a 3dlut for HDR, they just measure couple of colors and generate the 3dlut.
Let alone that LG has 1dlut and 3dlut in them, so I don't even understand what Stacy wants to experiment with: you would have to remove the 3dlut in the TV (1dlut can remain in place) and try to create one for madvr hdr output.
(But with recent nvidia drivers I guess even madTPG is screwed ...)

Yea I don't fully understand either. I was going to start playing with the idea of creating a 3DLUT in madvr for HDR to see what that looks like...just for fun. I doubt it'll be useful for us.

I think Stacy Spears wants to do an experiment with Roll-Off points set to 100% and see how it works with madVR TM ( possible with Device Control or Calman on C9 and newer models) :

https://imgur.com/9P1myeH.jpeg

Right so if you set the roll off higher than what madvr tone maps to, then in theory you have "pure" tone mapping with madvr as the tv wouldn't touch anything below the rolloff point. That's the theory. I'm curious how that plays out.

quietvoid

27th June 2021, 22:55

I don't see any flashes for now with your curves (FastBlur enabled here as well), maybe higher nits contents are affected?

Yes it's only for anything over DPL, it's pretty random for me though.

SamuriHL

27th June 2021, 23:31

I didn't see any last night but I only watched one movie and as we know I had a hard system lock in the middle of it so that really doesn't count for much. LOL

chros

28th June 2021, 01:21

Yes it's only for anything over DPL, it's pretty random for me though.
Cheers, maybe with the next build, I went back to b113 once more... due to neverending bugs...

SamuriHL

28th June 2021, 02:23

I'm trying to profile my HDR mode now. I.....don't think this is going to go very well. LOL! HDR is a mess on these screens. Which is what I've been told repeatedly. We'll see. I'm going to upload the 3DLUT to madvr's rec.2020 slot when I'm done and take a look. I don't know how long this will take though, and I may fall asleep before I can test it.

SamuriHL

28th June 2021, 03:42

JaylumX

28th June 2021, 04:17

That was a complete and total waste of my time. Won't be messing with this again. I created a 3DLUT based on profiling in HDR. When I select that LUT in the tonemap HDR with the 3DLUT option, it's outputting SDR. ummmmm, NO! Plus the fact that these panels DO NOT PROFILE PROPERLY FOR HDR. Period. There's no getting around that fact. I'm HIGHLY aggravated I went down this path and wasted 2 hours for nothing.

Have you tried calibrating (or generating) a 3DLUT in your display's brightest SDR mode and then loading it up in MadVR's Calibration menu as rec 2020 and then use tone map HDR using pixel shaders in the HDR menu (Do not use tone map HDR using external 3DLUT). That way it applies the 3DLUT form the rec 2020 part to the HDR image. This is on using MadVR's latest build - v134. The result, at least on my monitor is stunning. I just watch John Wick Chapter 2 and it was like watching a new movie. It looked good before but with the 3DLUT loaded, it is on a different level. I can't go back now.

You can also calibrate separately for SDR using rec 709 so that the SDR and HDR have separate 3DLUTS configurations.

SamuriHL

28th June 2021, 04:20

My c8 gets 400 nits max in sdr mode and that really pisses it off. It's a far cry from the 800 or so nits we get with hdr mode. My hope was that it'd tone mapping to hdr but that doesn't appear to be the case.

Edit... wait I re read what you posted and I think I somewhat understand now. You're saying under the display is calibrated settings rather than the hdr settings to load a rec2020 sdr 3dlut and then when hdr is displayed in bt2020 the 3dlut will be applied. Hmm that's interesting.

Sent from my SM-G998U1 using Tapatalk

JaylumX

28th June 2021, 05:04

SamuriHL

28th June 2021, 05:27

I'll try it sometime this week when I get a chance.

Sent from my SM-G998U1 using Tapatalk

quietvoid

28th June 2021, 14:07

In SDR that would imply getting BT2020 to DCI-P3 properly rendered on the display (in SDR) though, no?
I'm not sure if Native color would do that.

I just have a calibrated 1D LUT and an absolute colorimetric BT2020 to DCI-P3 3D LUT generated with DisplayCal, both uploaded to the TV.
Both contrast and colors look a lot better than factory for me.

Only issue is that for some reason black bars are randomly grey during playback.
Full field black isn't so I have no idea why that happens :(

SamuriHL

28th June 2021, 14:21

I have my sdr mode on the c8 calibrated to both rec709 and dci-p3 (using the rec2020 lut slot). So that's not an issue. The issue is that when playing hdr it doesn't use that rec2020 lut slot obviously. But if I can generate a madvr rec2020 lut and have it get used in hdr that would be wildly amazing I think. Not something I had even considered previously.

Sent from my SM-G998U1 using Tapatalk

ashlar42

28th June 2021, 14:41

Anybody with some guidelines for creating custom resolutions through CRU with LG OLEDs (CX or C1 should be identical, both 40Gbps HDMI 2.1)?

By opening CRU I see no trace of 2160p23 and 2160p59 resolutions, only 24, 25, 30, 50 and 60 (plus 100 and 120). But they are selectable from Windows and madVR and Kodi switch to them.

https://i.imgur.com/sPuJvxo.jpg

https://i.imgur.com/prouCFJ.jpg

https://i.imgur.com/L6PFjeA.jpg

madVR sees the 23Hz resolution

https://i.imgur.com/RimQetP.jpg

https://i.imgur.com/giLp32x.jpg

VBB

28th June 2021, 18:13

What are you trying to accomplish? Better than one frame drop/repeat per hour?

ashlar42

28th June 2021, 18:15

What are you trying to accomplish? Better than one frame drop/repeat per hour?Yes. Current timings provide one frame drop every 44.40 minutes.

VBB

28th June 2021, 18:19

ashlar42

28th June 2021, 18:33

Forgot which card you have, but most of us haven't bothered with custom modes since Nvidia adjusted their drivers a year and a half ago or so. Timings went from one drop/repeat every four minutes to a little over an hour.
RTX 3070. Is the interface for custom resolutions finally usable? I remember all sorts of problems in the past, which is why I went with CRU years ago.

VBB

28th June 2021, 18:44

The built-in madVR one should work, but like I said, many haven't bothered, because 1h5m is good enough. I've never used CRU. Let us know how it goes.

chros

28th June 2021, 18:59

I think Stacy Spears wants to do an experiment with Roll-Off points set to 100% and see how it works with madVR TM

Right so if you set the roll off higher than what madvr tone maps to, then in theory you have "pure" tone mapping with madvr as the tv wouldn't touch anything below the rolloff point. That's the theory. I'm curious how that plays out.
Oh, I see, yes, that can be the one. Although I haven't seen this in action, but I doubt there's any (huge?) visible gain doing this (unlike with "proper" LUTs for HDR10).

I just have a calibrated 1D LUT and an absolute colorimetric BT2020 to DCI-P3 3D LUT generated with DisplayCal, both uploaded to the TV.
Both contrast and colors look a lot better than factory for me.

Only issue is that for some reason black bars are randomly grey during playback.
Full field black isn't so I have no idea why that happens :(
I guess under randomly you mean with certain files and not during playback of a given file, right? If so, are you sure that those files are properly encoded/flagged?
As I said previously, I don't have this issue at all, but there are wrong files, so for those I have to adjust "source level" in madvr.

Have you tried calibrating (or generating) a 3DLUT in your display's brightest SDR mode and then loading it up in MadVR's Calibration menu as rec 2020 and then use tone map HDR using pixel shaders in the HDR menu (Do not use tone map HDR using external 3DLUT). That way it applies the 3DLUT form the rec 2020 part to the HDR image.
...
The result, at least on my monitor is stunning.

My c8 gets 400 nits max in sdr mode ...
Edit... wait I re read what you posted and I think I somewhat understand now. You're saying under the display is calibrated settings rather than the hdr settings to load a rec2020 sdr 3dlut and then when hdr is displayed in bt2020 the 3dlut will be applied. Hmm that's interesting.

I have my sdr mode on the c8 calibrated to both rec709 and dci-p3 (using the rec2020 lut slot). So that's not an issue. The issue is that when playing hdr it doesn't use that rec2020 lut slot obviously. But if I can generate a madvr rec2020 lut and have it get used in hdr that would be wildly amazing I think. Not something I had even considered previously.
This is way more complicated topic, I highly doubt that creating an SDR 3DLUT will be "accurate" for HDR in madvr.
Ask madsh's stake on this topic, but probably he didn't even think about this (just enabled it).

The problems are that:
- SDR mode on WOLEDs are different than HDR modes in terms of color accuracy (they drive the panel difefrently)
- in the processing chain of our LG TVs, they apply the built-in 1d/3dluts at the end of the chain, targeting gamma space 1d/3dluts
- with hdr output, madvr works in PQ space (not gamma space) -> so I guess a gamma sdr 3dlut won't be good in madvr for hdr10 content

If the above is true, I'm not even sure we can create a PQ 3dlut with DisplayCal for "normal" 3dlut slots.

But since we have this option, we can ask madshi, again, to try to disable the bt2020 flag with hdr output: this way the TV would only use it's native gamut and madvr could do the heavy lifting.

chros

28th June 2021, 19:07

RTX 3070. Is the interface for custom resolutions finally usable? I remember all sorts of problems in the past, which is why I went with CRU years ago.
I don't know whether it's usable now (depends on the driver/OS), but you can try my setting (https://forum.doom9.org/showthread.php?p=1868998#post1868998), maybe it will also work for you with new driver/OS/card (your original EDID values match mine).

quietvoid

28th June 2021, 19:09

I guess under randomly you mean with certain files and not during playback of a given file, right? If so, are you sure that those files are properly encoded/flagged?
As I said previously, I don't have this issue at all, but there are wrong files, so for those I have to adjust "source level" in madvr.

No, randomly in playback for any HDR10 video. It happens for HLG too. The bars would just become grey and black again after a bit, or sometimes never.
I've read some posts about "flickering black/grey" on AVS but I think it was more specific to Dolby Vision.

Levels are fine. I probably should factory reset and just upload the LUTs to see if it still happens.
Otherwise the only other thing is if I messed up anything in the service menu. I think I disabled ASBL and something else.

SamuriHL

28th June 2021, 19:17

Oh, I see, yes, that can be the one. Although I haven't seen this in action, but I doubt there's any (huge?) visible gain doing this (unlike with "proper" LUTs for HDR10).

Yea, I think that's why Stacey wants to experiment with it. It's an intriguing idea.

This is way more complicated topic, I highly doubt that creating an SDR 3DLUT will be "accurate" for HDR in madvr.
Ask madsh's stake on this topic, but probably he didn't even think about this (just enabled it).

The problems are that:
- SDR mode on WOLEDs are different than HDR modes in terms of color accuracy (they drive the panel difefrently)
- in the processing chain of our LG TVs, they apply the built-in 1d/3dluts at the end of the chain, targeting gamma space 1d/3dluts
- with hdr output, madvr works in PQ space (not gamma space) -> so I guess a gamma sdr 3dlut won't be good in madvr for hdr10 content

If the above is true, I'm not even sure we can create a PQ 3dlut with DisplayCal for "normal" 3dlut slots.

But since we have this option, we can ask madshi, again, to try to disable the bt2020 flag with hdr output: this way the TV would only use it's native gamut and madvr could do the heavy lifting.

I have to see what I can create with ColourSpace. I MIGHT be able to force it to create a PQ 3dlut that I can export and pull into madvr in the sdr rec.2020 slot. Obviously I'd need to create profiles to make sure it's not used in sdr content, but this at least opens up possibilities. Maybe. I'm not sure your idea for stripping the bt2020 flag for HDR will work. I can certainly try it myself and see what it'd do. I have the ability to strip it already. We'll see how much playing around I get to do tonight.

ashlar42

28th June 2021, 20:14

The built-in madVR one should work, but like I said, many haven't bothered, because 1h5m is good enough. I've never used CRU. Let us know how it goes.
Nope, it doesn't. But I can do it by "hand". Problem is using the right tool. And currently I don't know how to correctly approach CRU with this new panel. I have written in the support forum, we'll see.

I don't know whether it's usable now (depends on the driver/OS), but you can try my setting (https://forum.doom9.org/showthread.php?p=1868998#post1868998), maybe it will also work for you with new driver/OS/card (your original EDID values match mine).
This is what I was planning to do, starting from there and calculating the offset according to my sync values. I will update if I come to something. Thanks!

EDIT: Nvidia control panel did not let me add a custom 2160p23 resolution. CRU... I tried adding a resolution with correctly calculated values, but madVR reports a very different refresh rate compared to the one calculated.

Aiming for 23.976399 (calculated with Pixel Clock Calculator, on monitortests.com), I got 23.977xxx. I'm probably too old for this s$£%, but I'm so fed up with custom resolutions being some sort of black magic ritual. And no, I don't like three repeated frames in a 2h30m movie.

chros

28th June 2021, 22:39

I'm not sure your idea for stripping the bt2020 flag for HDR will work. I can certainly try it myself and see what it'd do. I have the ability to strip it already. We'll see how much playing around I get to do tonight.
Really?! :) Then please can you check it? Take a look whether LG still reports the bt.2020 flag at the top of the info panel with HDR10.

SamuriHL

28th June 2021, 23:27

I can but it's not going to be today. I had a major work crisis today and just finished so it's going to definitely be tomorrow before I touch anything. But yea, I can remove the bt.2020 flag with the Integral 2 and see what it does. At least I think I can. I'll find out tomorrow.

SamuriHL

29th June 2021, 02:35

quietvoid

29th June 2021, 03:01

@chros So I managed to find a scene where black bars are consistently elevated if you want to test: https://0x0.st/-p6H.mkv
I only get raised bars before the full black cut.

Tested all settings + service menu and I can't get rid of it.
As soon as I upload my 1D LUT, it happens.

chros

29th June 2021, 07:35

I just have a calibrated 1D LUT and an absolute colorimetric BT2020 to DCI-P3 3D LUT generated with DisplayCal, both uploaded to the TV.
Both contrast and colors look a lot better than factory for me.

Only issue is that for some reason black bars are randomly grey during playback.
Full field black isn't so I have no idea why that happens :(
Now as you mentioned, I saw this (at least) once: it was with a bravia core content (don't remember which one), but most of those titles are full range (!), not limited, so I thought it's the source (and I haven't checked in other factory preset).

@chros So I managed to find a scene where black bars are consistently elevated if you want to test: https://0x0.st/-p6H.mkv
I only get raised bars before the full black cut.

Tested all settings + service menu and I can't get rid of it.
As soon as I upload my 1D LUT, it happens.
That's weird, for me this file is fine, no raised black bars before black screen, I tried:
- a preset with custom 1dlut + factory 3dlut
- a preset with both custom 1d/3dlut
- madvr (b113) and TV set to limited or full range signal
- ASBL is disabled in Service menu here as well (there's no GSR OFF option for me with the old firmware I use)

But you know guys, you have newer firmware in your C8 compared to my B8, LG did some change in newer firmwares to them due to the chrominance overshoot fix, so who knows.
Maybe @SamuriHL can try this for you, if he also has 1dlut uploaded for hdr10.

Is factory preset fine with this, e.g. Cinema Home?

chros

29th June 2021, 07:46

Ok, so I was able to give that a quick test. I can indeed strip the bt.2020 flag out of the HDR signal. TV reports HDR and enables HDR mode but does NOT report bt.2020. So that's fun. :D It's a real pain in the ass to set up on the Integral 2, too, however. Nonetheless it does seem to do what you want.
Thank You! Exactly what we were wondering about.
Questions:
- do you remember any picture mode change (new hdmi handshake) when you enabled/disabled the flag? (there's none using SDR and enabling/disabling it via madvr)
- did you see any color difference with and without it? (I don't see any using SDR)

Now I'll write a request about this, again, to madshi, if he can / is willing to implement it ... Come over there and support it, otherwise nothing will happen :)

chros

29th June 2021, 10:40

Oh, and aiopylgtv added methods (https://github.com/bendavid/aiopylgtv/commit/711abda131203ea9dc3076736f96b3316af7a5bb) to switch picture presets (with this hack (https://github.com/bendavid/aiopylgtv/commit/e2607dd8bb51b057e6197e0c6b5f9dfe2310d862) :D), this is really useful not to use UI elements on screen that can modify image with HDR10/DoVi content:
aiopylgtvcommand 192.168.1.78 set_current_picture_mode expert2

SamuriHL

29th June 2021, 13:47

But you know guys, you have newer firmware in your C8 compared to my B8, LG did some change in newer firmwares to them due to the chrominance overshoot fix, so who knows.
Maybe @SamuriHL can try this for you, if he also has 1dlut uploaded for hdr10.

Is factory preset fine with this, e.g. Cinema Home?

I don't ever touch factory 1dlut for this very reason. I have mentioned it before that there seems to be some LG bug with 1dlut causing raised blacks. It's been discussed in both the ColourSpace and Calman threads numerous times. Although Calman claims to have worked around it with something they call a "shadow detail tweaker" whatever the hell that means. With ColourSpace we don't bother with 1dlut and just create a 3dlut, but, not for HDR. So I can't really help here. I can tell you I don't have any raised black issues with the way I've calibrated.

SamuriHL

29th June 2021, 13:55

Thank You! Exactly what we were wondering about.
Questions:
- do you remember any picture mode change (new hdmi handshake) when you enabled/disabled the flag? (there's none using SDR and enabling/disabling it via madvr)
- did you see any color difference with and without it? (I don't see any using SDR)

Now I'll write a request about this, again, to madshi, if he can / is willing to implement it ... Come over there and support it, otherwise nothing will happen :)

That's a difficult question to answer because using the Integral2 to change the flag it most definitely did an HDMI handshake because of the way I have to do it. So my test isn't valid to answer that question in the way you're looking for.

The answer to question 2 is also difficult. There isn't an option to just strip bt.2020. You have to force a container change so in my case there was a color difference but only because of how I modified the signal by forcing it into an RGB full container. Now if I had output that from my HTPC from the start, it'd probably work without color issues.

I was only really looking to see if it really stripped the bt.2020 flag and not focused on much else. But it definitely seems to do what you're looking for. That being said, madshi posted a ton of information on AVS that I don't fully follow. I kind of do. The way you calibrate HDR on these tv's is in gamma 2.2. Internally the TV converts gamma 2.2 to PQ, much like what madshi was saying. Calman for sure uses 2.2 to calibrate HDR and I'm pretty sure the C9 and above users calibrate the same way in ColourSpace. But these LG panels assume 2.2 and likely store the LUT internally as 2.2, although that's a guess on my part and could be wildly wrong. Maybe it's converting the 2.2 to PQ when it stores the lut, I don't really know for sure. This discussion is WAY above my pay grade. This is not an area I have any real knowledge in. Only what I've read.

SamuriHL

29th June 2021, 14:04

I'm still following the discussion between omarank and madshi with great interest. omarank is explaining it very well. As I said, the calibration software generates the lut in 2.2 and the display internally converts that to PQ. He's telling madshi that for madvr to convert that to PQ to send to the lut is wrong. And I think I agree with that assessment. This is exciting if madshi will be convinced because if he makes that change, or at least allows a test of that theory, this could provide us with a VERY kuel option!!!

quietvoid

29th June 2021, 14:19

But you know guys, you have newer firmware in your C8 compared to my B8, LG did some change in newer firmwares to them due to the chrominance overshoot fix, so who knows.

Is factory preset fine with this, e.g. Cinema Home?

Yea it might be related to the firmware, since it's mostly happening when the actual video adjacent to the bars is near black.
Factory preset Cinema restores the bars back to black.

I tried reencoding and cropping the bars, but the result is the same so it's on the TV.

Oh well, I guess I'll have to live with it for the better PQ tracking.

I don't ever touch factory 1dlut for this very reason. I have mentioned it before that there seems to be some LG bug with 1dlut causing raised blacks. It's been discussed in both the ColourSpace and Calman threads numerous times. Although Calman claims to have worked around it with something they call a "shadow detail tweaker" whatever the hell that means. With ColourSpace we don't bother with 1dlut and just create a 3dlut, but, not for HDR. So I can't really help here. I can tell you I don't have any raised black issues with the way I've calibrated.

Yea there are multiple mentions of raised blacks with 1D LUTs.

Calman's shadow detail seems to be about altering the 1D LUT values to increase brightness in shadows, but that has nothing to do with the black point.
It might be useful for daytime watching (ala 2.2 gamma instead of 2.4 during the day).

I have no idea how to create a 3D LUT that would include gamma fixes for HDR. So a 1D LUT Is still unavoidable for me.

SamuriHL

29th June 2021, 14:27

Yea there are multiple mentions of raised blacks with 1D LUTs.

Calman's shadow detail seems to be about altering the 1D LUT values to increase brightness in shadows, but that has nothing to do with the black point.
It might be useful for daytime watching (ala 2.2 gamma instead of 2.4 during the day).

Ah, ok. I don't have a current license for it so haven't really looked too much into what it's doing. That makes sense though.

I have no idea how to create a 3D LUT that would include gamma fixes for HDR. So a 1D LUT Is still unavoidable for me.

You can't. The only 3dlut you can create for HDR on these panels is a matrix lut for the primary colors. It's not useful. All the grayscale stuff has to be done in the 1dlut which puts you in a catch 22.

ashlar42

29th June 2021, 14:29

quietvoid

29th June 2021, 15:03

The package setup creates an alias for aiopylgtvcommand but it's still all based on Python.

chros

30th June 2021, 09:36

Is there a command line interface somewhere? Or one needs to use python?
You can easily install it on Windows:
A. Install python
- https://www.python.org/downloads/release/python-382/
- - python-3.8.2-amd64
- check Install launcher for all users
- check Add Python 3.8 to path
- select custom install: select the path, e.g. D:\Progs
B. Install aiopylgtv
- download: https://github.com/bendavid/aiopylgtv/archive/master.zip
- unzip it: d:\Progs\aiopylgtv\
cd d:\Progs\aiopylgtv\
pip install --upgrade .
- to try it out (TV must be on!):
aiopylgtvcommand 192.168.1.78 info_button
- - it will create the .aiopylgtv.sqlite db at first run
- - although it can be run from any directory, but .aiopylgtv.sqlite has to be copied into that directory!!!

chros

30th June 2021, 09:42

The answer to question 2 is also difficult. There isn't an option to just strip bt.2020. You have to force a container change so in my case there was a color difference but only because of how I modified the signal by forcing it into an RGB full container. Now if I had output that from my HTPC from the start, it'd probably work without color issues.
I see, thanks for checking.

Yea it might be related to the firmware, since it's mostly happening when the actual video adjacent to the bars is near black.
Factory preset Cinema restores the bars back to black.

I tried reencoding and cropping the bars, but the result is the same so it's on the TV.

Bloody LG, always screws up something ... Such an incompetent bunch of fellows :)

chros

30th June 2021, 09:52

@quietvoid, about FireStick and DoVi: people in the new Sony threads (A80J, A90J) complained about that certain external devices (Firestick, Roku, etc., but Shield is ok) produce darker image while streaming DoVi content on them compared to the built-in apps on the TV. They came to the conclusion that probably these devices utilize player-led DoVi and not TV-led.

I also use FireStick for streaming and with Kodi (with your patches), so after reading the above, I started to compare Netflix (ATV+ and Disney+ apps are not available with the old firmware) DoVi content with the built-in app and from Firestick (using the same DoVi picture preset): I also noticed that the image is just a bit darker using FireStick. It's not a night and day difference as with Sonys but it's there.
Plus we know, that the Oppo has to be set to TV-led manually, because Auto triggers player-led.

What do you think, have you noticed anything about this? Unfortunately I couldn't test this (as I did with the Oppo DoVi upconverting feature), since I don't have proper test content for this.

SamuriHL

30th June 2021, 11:16

I agree with your LG assessment. Sigh

Sent from my SM-G998U1 using Tapatalk

quietvoid

30th June 2021, 14:02

chros

30th June 2021, 18:19

I've compared a couple times but couldn't really notice a brightness difference.
Yes, it's not (much/any?) brightness diff with our LG but as I said earlier LG disables the same image processing with player-led DoVi just as in PC mode, resulting in posterization in dark scenes (and who knows what else that we haven't noticed yet).

These days I stream Netflix using Kodi on the Firestick since it's a lot more convenient, and I don't have to deal with adaptive quality changes as opposed to the official apps.
Interesting, can you point out what/how needs to be installed? Thanks

quietvoid

30th June 2021, 20:07

Interesting, can you point out what/how needs to be installed? Thanks

The Netflix add-on is required: https://github.com/CastagnaIT/plugin.video.netflix
Properly installed you should be able to get automatic highest quality, as well as the option to change both audio/video streams.

There is a FAQ here: https://github.com/CastagnaIT/plugin.video.netflix/wiki/FAQ-%28Audio%2C-Video%2C-Subtitle%2C-Other%29#my-android-device-supports-4k-but-does-not-playback-at-4k

huhn

1st July 2021, 06:53

how terrible is a end of 2020 CX6 in game/PC mode?

chros

1st July 2021, 11:28

I think Asmodian has a CX, what we know, that:
- they finally fixed banding in PC mode
- but posterization is still there in dark scenes
- we don't know anything about potential color issues with hdr10 (that plagued 2018 LGs) (this one also happens in normal mode)

And who knows what else they screwed up :) (banding, grid, etc)
Do you want to get one?

@quietvoid: thanks, I'll try it out.

huhn

1st July 2021, 11:39

after 5 month of torture i can finally return my xh90 so yes i need a new TV and there are no new TVs available it's like the GPU market.

so i found a 55cx6 for ~1100 and if it is one of the better oled for PC use with 120 HZ and VRR than that's a very not bad alternative.
i would prefer a 65 one but tehy cost a lot more for only a small size upgrade so that just feels bad and the last year cx is not available here in 65 anymore.

i wanted a bad samsung 65q70t but they are not available anymore and the q70a is an abomination that should be burned with fire.

chros

1st July 2021, 19:33

after 5 month of torture i can finally return my xh90
What was the issue with xh90?

q70a is an abomination that should be burned with fire
:D I'd never buy LCD ever again ...

a 55cx6 for ~1100 and if it is one of the better oled for PC use with 120 HZ and VRR
Price is not bad (as it was 2 years ago), but be prepared that no TV is perfect, as we all know. Buy from a place with big return window (at least 2 weeks), if there's anything with it (banding, tinting, grid, etc) just return the whole set (not just panel replacement by LG) and try another one!

I understand that for us who already has a working TV, it's not that urgent to buy something now. That's why most of us haven't bought a CX (LG dumbed it down for whatever reason), this year's C1 is dumbed down again (no stand for G1).
What we understand now is, that next year (again) "new" panels will be introduced with modified pixel structure, so maybe 2022 LG models will be the way to go, and hopefully they return to the 1 high end line for C2/G2/etc. We will see.
But personally I refuse to pay the opening price, so that means almost 2 years from now for me, if (!) all will be good with those models.

1 last note about LGs: they have a very bad history with firmware updates, so only upgrade the firmware if you desperately need a bugfix or new feature :)

If you go on the Oled route, feel free to ask us anything in the future :)

huhn

1st July 2021, 20:01

xh90 is literally a scam.
it's advertised with 4K 120 where it downscales to 1080 and than up again using bilinear. lots of TV have this problem this year because they use the same low end mediatek chip.
even through is advertised it can't do VRR at all and the TV is a year old.

OLED are generally not for me i manage to burn in an IPS TV i don't change my self for a device.

beware the CX6 i ordered is a European version from the end of 2020 so it could be quite different which i don't like it only has 1 tuner for example really interest what Asmodian has to say about this device.

chros

2nd July 2021, 11:12

xh90 is literally a scam.
Yeah, I heard the same about 900H, and I'd put a large sum on the same will happen with this years A80J/A90J Oleds as well. :)

the CX6 i ordered is a European version from the end of 2020 so it could be quite different
I don't think they are any different in terms of image functionality and potential bugs :)

quietvoid

2nd July 2021, 16:05

I see you guys talking about 3D LUTs but I don't understand what more it would bring for HDR. :o
Never used LUTs in madVR but I have them uploaded to the TV already.

chros

2nd July 2021, 16:28

HDR10 has the color issues due to bt2020 3dlut processing on 2018 LG Oleds (we don't know how newer models behave). If you remove the bt2020 3dlut, the issue is gone, but obviously you get desaturated colors. That's where the new madvr feature comes in.
E. g. that's why jk82 never watched hdr10 content on his C8!
(But if you only have 700 hours in your 2 years old TV then probably you missed this :))

As I said there (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60852164), it even fixed weird colors in PC mode (that are not present in normal mode), banding and near black posterization are still there. (This is how 3dluts work (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60844190) in our LG Oleds)

Anyway, if we manage to create an acceptable 3dlut for this, hdr10 watching will be enjoyable again :)

quietvoid

2nd July 2021, 16:32

Hm, I removed the factory 3D LUT but replaced it with the synthetic one, and increased Color to 55 on my HDR10 settings.
Can't see anything desaturated (or at least noticeably so).

Would be nice to be able to do an actual color calibration, but I'm not sure why it would be better to use it with madVR compared to uploading into the TV.

SamuriHL

2nd July 2021, 17:50

The issue is you can't upload a proper 3D LUT for HDR to the C8. It just doesn't work. So the idea is if you can get rid of the internal 3D LUT on the C8 for HDR and then create a proper one for madvr and let it correct before sending it to the C8, we will get far better and more accurate color representation than what can be done natively on the panel. So if madshi implements the option chros asked for, we'll be able to create a color accurate 3D LUT for madvr and bypass the stupid s*** the C8 is doing. That is indeed quite exciting.

EDIT: Of course I'll have to figure out how to replicate what you did in displaycal, chros. I'd like to use ColourSpace to replicate your success.

huhn

2nd July 2021, 20:13

Yeah, I heard the same about 900H, and I'd put a large sum on the same will happen with this years A80J/A90J Oleds as well. :)
they are the same just us name and eu name.
the xj90 has full reso in sdr and is supposed to have the same issue in HDR they don't ship with VRR no soiny TV up to this day has VRR.

I don't think they are any different in terms of image functionality and potential bugs :)

it's supposed to be released at the 20.10.20. i will not be happy anyway it's part of been an TV owner

j82k

2nd July 2021, 21:57

So, 3 days ago I got a LG C1. One of the main reasons for the upgrade was the messed up HDR mode of the C8. Other than for testing I never watched HDR on that TV for the last year or so. I used madVR to tonemap to SDR, which means ~340 nits.

And now suddenly we have this 'fix' that chros discovered. I just tried it with a synthetic madVR 3LUT (BT2020 G2.2 to P3-D65 G2.2) and put that into the BT2020 slot in madVR. For the C8 I used my manually calibrated 1DLUT + unity 3DLUT.

Checked all the scenes I remembered that show the cyan/magenta issue and they all looked okay, usually watching these makes me wanna puke. :)

I have now put the C8 and C1 side by side for comparison. The C1 is playing files from usb.
In this direct comparison the C8 still shows a bit more saturation than the C1 in problematic dark scenes but it doesn't show these ugly oversatrurated color transitions anymore.
Brighter scenes look very similar on the two TVs, so I guess that means that this method retains accurate EOTF tracking and saturation.

Now I'm strongly considering sending the C1 back. BFI is a huge disappointment, near-black overshoot still exists and I think I actually prefer the original weaker dithering of the C8 which is what you'll get when uploading a 1DLUT or if you stayed on an old firmware.

If only this was discovered a week ago, it could've saved me a lot of hassle. :)

SamuriHL

2nd July 2021, 22:45

Holy s***! If that's not a testament to this I don't know what is!!!!!!!!

chros

3rd July 2021, 11:58

Hm, I removed the factory 3D LUT but replaced it with the synthetic one, and increased Color to 55 on my HDR10 settings. Can't see anything desaturated (or at least noticeably so).
Of course, you modified Color (cheater! :) ) and you still have 3dlut in place (the synthetic one), remove that one by uploading a unity one:
aiopylgtvcommand 192.168.1.78 upload_3d_lut_bt2020 hdr_cinema

Would be nice to be able to do an actual color calibration, but I'm not sure why it would be better to use it with madVR compared to uploading into the TV.
What baffles me that apart from only couple of us no one was complaining about this stupid hdr10 color issues with 2018 models (not just here, but anywhere)!
Test it with The Witcher sample (https://forum.doom9.org/showthread.php?p=1925528#post1925528) (hdr10 vs DoVi) :)

we will get far better and more accurate color representation
I don't think we can talk about color "accuracy" with HDR10 at all :) Let's not call it calibration when they measure only 3 colors in 2021! :D

So if madshi implements the option chros asked for, we'll be able to create a color accurate 3D LUT for madvr and bypass the stupid s*** the C8 is doing.
Unfortunately, it still won't be color "accurate" with 2018 models: as omarank pointed out, we can't disable TV's tonemapping in PQ space in any way. (But I don't think this would be a big issue, still better than the washed out colors in calibration mode).

EDIT: Of course I'll have to figure out how to replicate what you did in displaycal, chros. I'd like to use ColourSpace to replicate your success.
Beware that CS (at least a year ago) had to do some extra step during 3dlut creation for madvr that completely screwed up colors (I only talk about simple SDR 3dluts). At that time Manni was really pissed off about them, not implementing it properly (CS works in full range internally, madvr requires limited signal).

So, if that's still the case, then just leave it, and do it with displaycal.
As I said there (https://www.avsforum.com/threads/madvr-argyllcms.1471169/post-60851944) I had to raise gamma to (actually) 2.11 (!) to do not get crushed shadow details (obviously normal scenes are way bright now), but maybe enabling "black point compensation" with curves+matrix profile will help it.
The good thing about it, that profiling is pretty fast measuring only 34 patches (grayscale+RGB), so lot variations can be tried out.

chros

3rd July 2021, 12:24

And now suddenly we have this 'fix' that chros discovered.
Actually You were the one who started to suspect bt2020 3dlut processing being the culprit at that time you reported (https://forum.doom9.org/showthread.php?p=1890535#post1890535) this (more than 1 and a half a year ago :) ). Of course at that time we didn't have the tools to verify this, and we just saw this working when we started to play with 1dluts (half a year ago).
I just put the pieces together :)

I just tried it with a synthetic madVR 3LUT (BT2020 G2.2 to P3-D65 G2.2) and put that into the BT2020 slot in madVR. For the C8 I used my manually calibrated 1DLUT + unity 3DLUT.
Can you please post screenshots of the 3 window (2 icm profile and 1 3dlut window) with settings you used? (e.g. onto imgbox)

I'm asking this, because every setting can make an impact (top nits, gamma on icm and 3dlut windows, rendering intent / etc).
I tried out at least 20 (!) different synthetic 3dluts (with same 1dlut uploaded into 2 presets) but none of them came close to the factory 3dlut (not that it has to be the etalon, but that's the only thing we can compare to).

Checked all the scenes I remembered that show the cyan/magenta issue and they all looked okay, usually watching these makes me wanna puke. :)
:D I know, I can't stand it either, that's why I started to use DoVi for everything I could.

I have now put the C8 and C1 side by side for comparison. The C1 is playing files from usb.
In this direct comparison the C8 still shows a bit more saturation than the C1 in problematic dark scenes but it doesn't show these ugly oversatrurated color transitions anymore.
Brighter scenes look very similar on the two TVs, so I guess that means that this method retains accurate EOTF tracking and saturation.
Nice, amazing that you have side-by-side setup now! :)
That's why I'd like to try out with synthetic 3dlut with Your settings to see how it goes.

Not sure you've seen this (https://forum.doom9.org/showthread.php?p=1946431#post1946431), it's just amazing!
Create separate batch files (bat/cmd) for them, and you can use my tool (https://github.com/chros73/madvr-js-remote) or SirMaster's (https://github.com/nicko88/HTPCRemote) to easily toggle between them!
I have setup mine to have 4 buttons for: 2 for madvr (passthrough, DTM) and 2 for LG (HDR Cinema, HDR Technicolor). So just pushing 2 buttons you can easily compare the images :)

BFI is a huge disappointment, near-black overshoot still exists
Why BFI is disappointing? Causing more near-black overshoot?

I think I actually prefer the original weaker dithering of the C8 which is what you'll get when uploading a 1DLUT or if you stayed on an old firmware.
I agree, what's the point of having it when it doesn't fix sh1t :D

chros

3rd July 2021, 12:43

Now I'm strongly considering sending the C1 back. ... If only this was discovered a week ago, it could've saved me a lot of hassle. :)
Not necessarly, C1 can be still a significant improvement over the C8. Can you test these (after the break-in period, after 50/100 hours)? Especially if you have the 2 TVs side by side :)

1. banding should be largely reduced in HDR10 PC mode, test it with Mehanik's HDR10 test patterns (03. Grayscale\02. Ramps\01. Diagonal Ramp 400nit.mp4) (or any other clip that produce this on C8)
2. what about near black posterization in PC mode? (last part of this post (https://forum.doom9.org/showthread.php?p=1883108#post1883108) for sample)
3. does DoVi have raised black and still crush near black? (you can use DoVi samples from here (https://forum.doom9.org/showthread.php?p=1931241#post1931241) and this one from Stacy (https://www.avsforum.com/threads/spears-munsil-uhd-hdr-benchmark-disc-discussion.3075780/page-43#post-59634582) (but that dual track, dual layer, so I can only play it with Oppo), more info (https://forum.doom9.org/showthread.php?p=1932220#post1932220)).

Please do the PC tests with RGB Full 12bit output in nvidia panel + madvr set to 10bit or higher.

j82k

3rd July 2021, 14:00

Can you please post screenshots of the 3 window (2 icm profile and 1 3dlut window) with settings you used? (e.g. onto imgbox)

I'm asking this, because every setting can make an impact (top nits, gamma on icm and 3dlut windows, rendering intent / etc).
I tried out at least 20 (!) different synthetic 3dluts (with same 1dlut uploaded into 2 presets) but none of them came close to the factory 3dlut (not that it has to be the etalon, but that's the only thing we can compare to).

I just used the DCI-P3 D65 preset and changed gamma to 2.2 and in the 3D Lut Maker the standard BT2020 profile with gamma 2.2. White Level doesn't matter, I tried 100 nits and 10K nits, no difference.

Rendering intent absolute colorimeric for now but I'll probably try the other ones.

Why are you even trying to create a matrix LUT in displayCal? Couldn't you just enter your measured primaries into the ICC profile creator? That would get you the same result but without messing up the EOTF which you can do much better with the 1DLUT.

https://images2.imgbox.com/2d/34/fJX3KS7j_o.jpg (https://imgbox.com/fJX3KS7j)

Why BFI is disappointing? Causing more near-black overshoot?

Yes that and it also messes up the near-black range. I didn't even try to calibrate the TV with BFI as the increased overshoot is already reason enough to not use it.
Also I don't find that it actually helps with camera panning scenes with 24P content, it still looks like a stuttery mess.

Not necessarly, C1 can be still a significant improvement over the C8. Can you test these (after the break-in period, after 50/100 hours)? Especially if you have the 2 TVs side by side :)

1. banding should be largely reduced in HDR10 PC mode, test it with Mehanik's HDR10 test patterns (03. Grayscale\02. Ramps\01. Diagonal Ramp 400nit.mp4) (or any other clip that produce this on C8)
2. what about near black posterization in PC mode? (last part of this post (https://forum.doom9.org/showthread.php?p=1883108#post1883108) for sample)
3. does DoVi have raised black and still crush near black? (you can use DoVi samples from here (https://forum.doom9.org/showthread.php?p=1931241#post1931241) and this one from Stacy (https://www.avsforum.com/threads/spears-munsil-uhd-hdr-benchmark-disc-discussion.3075780/page-43#post-59634582) (but that dual track, dual layer, so I can only play it with Oppo), more info (https://forum.doom9.org/showthread.php?p=1932220#post1932220)).

Please do the PC tests with RGB Full 12bit output in nvidia panel + madvr set to 10bit or higher.

I tried some of the scenes that show heavy posterization/color banding on the C8 in pc-mode. On the C1 they do look identical in pc-mode and non-pc mode. I even took some pictures for comparison but there really isn't anything to see as they look the same. That was with 8-bit though, haven't tried 12-bit.

So Pc-mode seems to be fixed on the C1 but for me this isn't that important as I use the TV 99% for movies/TV series and from viewing distance I don't see the benefit of 4:4:4 chroma.

Can't say anything about dolby vision, the only thing I tried was a dual layer mp4 remux that plays fine as dolby vision using the internal media player on the C8 but it only plays as HDR on the C1.

By the way the C1 is already packed up, waiting to be returned. I noticed some reverse vignetting on the left side and that bothered me more and more. The C8 has some too but it's mostly at the bottom which is much better than having it on the side.

chros

3rd July 2021, 19:59

Yes that and it also messes up the near-black range. I didn't even try to calibrate the TV with BFI as the increased overshoot is already reason enough to not use it.
Also I don't find that it actually helps with camera panning scenes with 24P content, it still looks like a stuttery mess.
Interesting, thanks. That's just stunning, that I haven't heard/read about this anywhere, not from a single "reviewer" ... I'm sorry, influencer ... :)

I tried some of the scenes that show heavy posterization/color banding on the C8 in pc-mode. On the C1 they do look identical in pc-mode and non-pc mode. ... That was with 8-bit though, haven't tried 12-bit.
It should 8/12 bit shouldn't do a big difference though, but who knows.
That's good to hear you haven't noticed anything outstanding, maybe (!) they fixed it. I asked guys with CX on avs a year ago with the posterization sample I linked and they confirmed that it was still there with CX. I even told this to Stacy on bluray.com forum (he stated that PC mode is perfect now with CX), never heard back anything :)
There are CX owners here as well, I don't recall anyone tested it either. Myabe @huhn will be the first? :)

Pc-mode ... I use the TV 99% for movies/TV series and from viewing distance I don't see the benefit of 4:4:4 chroma.
That's strange, I do, easily! I can spot PC vs non-PC in a blind test easily from ~3m (not like just guessing with Lanczos vs NGU upscaling :) ). It's just way more sharper.

Can't say anything about dolby vision, the only thing I tried was a dual layer mp4 remux that plays fine as dolby vision using the internal media player on the C8 but it only plays as HDR on the C1.
Thanks, that confirms the lack of dual DoVi decoder chip in C1 as well, just as with CX (last models were the 2019 ones). Why LG removed it??? :)
@huhn, DoVi blacks should be fixed on CX (at least at the time with the firmware when Vincent created a video about it), you can test it with the BlackPludge pattern above (unfortunately Stacy's pattern needs an external bluray player).

By the way the C1 is already packed up, waiting to be returned.
:D That was quick! Sorry to hear that, at least you spare some money :)

I just used the DCI-P3 D65 preset and changed gamma to 2.2 and in the 3D Lut Maker the standard BT2020 profile with gamma 2.2. Rendering intent absolute colorimeric for now but I'll probably try the other ones.
Thanks for the screenshots, now I can create exactly the same.

White Level doesn't matter, I tried 100 nits and 10K nits, no difference.
It did for me when I uploaded those into the TV, but created different profiles for bt2020 icm as well. Let's see if it does for me with madvr.

Why are you even trying to create a matrix LUT in displayCal? Couldn't you just enter your measured primaries into the ICC profile creator? That would get you the same result but without messing up the EOTF which you can do much better with the 1DLUT.
Good question and point, dunno. We have to experiment with this and comparing to other preset what looks good or not.

Maybe I'll ask madshi about the synthetic profiles, what's his take on it, he knows the best how madvr works internally.

huhn

3rd July 2021, 21:10

is it normal that the colors shift majorly even when you are 1 degree of angle? while it never ever get's as extreme as a VA panel is costly there even moving my head very slightly changes where the image is more red...

i'm trying to do a general setup for PC mode and madVR but wow these devices are terrible out of the box... i have to give sony here a clear advantage graphic or game mode and white point done that's all.

chros

3rd July 2021, 21:36

is it normal that the colors shift majorly even when you are 1 degree of angle?
Yes, something to do with the coating of the screen I think.

general setup for PC mode and madVR but wow these devices are terrible out of the box
Yes, in PC mode LG lockes gamut to "Wide" mode (not "Auto", that would mean bt709) which means the panel native DCI-P3 colorspace, so all bt709 SDR content will be oversaturated without having a calibration. :)

huhn

3rd July 2021, 21:46

genius...

and it looks like your gamut notice is correct while i have 3 option they all look 100 % the same.

so do i get this correctly that i can't have bt 709 SDR and bt 2020(DCI-P3 D65) at the same time? and i only get bt 709 after doing a full calibration(if that's even possible)?

just a reminder i payed for this device right? yes i did damn.

j82k

3rd July 2021, 22:07

It should 8/12 bit shouldn't do a big difference though, but who knows.
That's good to hear you haven't noticed anything outstanding, maybe (!) they fixed it. I asked guys with CX on avs a year ago with the posterization sample I linked and they confirmed that it was still there with CX. I even told this to Stacy on bluray.com forum (he stated that PC mode is perfect now with CX), never heard back anything :)
There are CX owners here as well, I don't recall anyone tested it either. Myabe @huhn will be the first? :)

Here is a comparison of PC Mode. There might be a tiny difference in color but I couldn't see any posterization or color banding issues caused by PC-mode with the C1 in this or other scenes.
The C8 on the other hand.... :eek:

https://imgur.com/a/oRxVUC5

chros

3rd July 2021, 22:24

https://imgur.com/a/oRxVUC5
Looks awesome, thanks, maybe they did fix it :) Now I will be curious of @huhn's test.

it looks like your gamut notice is correct while i have 3 option they all look 100 % the same.
So, you say that you can still select them in PC mode? That's weird, maybe an UI bug, that option should be greyed out and "Wide" selected.

so do i get this correctly that i can't have bt 709 SDR and bt 2020(DCI-P3 D65) at the same time?
What do you mean about "the same time"? (I don't understand.) And what do you mean about bt2020 here? Are we still talking about SDR?

and i only get bt 709 after doing a full calibration(if that's even possible)?
That's correct using bt709 SDR in PC mode: you can create a 3dlut for the TV or madvr to correct this.

just a reminder i payed for this device right? yes i did damn.
:D Yes, but it's not a big deal, you can correct this with calibration. Do you have an i1DiplayPro colorimeter?
Believe me, if everything else is working as it should (that you want to use), you hit the jackpot! :)

SamuriHL

4th July 2021, 04:42

I think what huhn is asking, and I could very well be misinterpreting here, is if you can use one mode for both rec.709 and bt.2020 (or in our case in reality dci-p3). You can. Each display mode has 2 3D LUT slots you can populate. You have to have the right software to calibrate them both at the same time but it's doable. I'm guessing DisplayCal can do it, too I just don't have experience with it. I have my bright mode calibrated with both rec.709 and bt.2020 LUTs. It works great.

If that's not what you were saying huhn, then I'm not sure.

huhn

4th July 2021, 07:02

So, you say that you can still select them in PC mode? That's weird, maybe an UI bug, that option should be greyed out and "Wide" selected.
it's worse/better...
it only work at all in PC mode it doesn't do anything outside of it
it's always doing nothing outside of PC mode and it always available in all modes i tested. if it is in instant response mode (this should trigger PC mode) freesync premium, isf dark cinema with and without deep cooler.

every still looks really really of even at auto which has the lowest saturation so it should bt BT 709. my xh90 was nearly calibrated out of the box my 3D lut does nearly anything i still have that one. sony only had an 11 point ire calibration and that was more than enough.
What do you mean about "the same time"? (I don't understand.) And what do you mean about bt2020 here? Are we still talking about SDR?
i want later used send bt 2020 option in madVR for HDR -> SDR tone mapping but in native without ever using the remote like every TV before could do.

That's correct using bt709 SDR in PC mode: you can create a 3dlut for the TV or madvr to correct this.
i heard they support it but i could not yet find it only 22 point.

:D Yes, but it's not a big deal, you can correct this with calibration. Do you have an i1DiplayPro colorimeter?
Believe me, if everything else is working as it should (that you want to use), you hit the jackpot! :)

the coating is a total deal breaker it turns a ~20X20 cm part on the screen red that should have a blue tint.

and yes i have a colorimeter and i will use it.

BTW. banding in PC mode isn't bad here it doesn't even challenge the xh90 which was perfect in 8 bit mode but it isn't so bad that i should ever notice it with real content. i only tested 8 bit right now i now regret not getting a HDMI 2.1 card just because my last screen 120 HZ was a lie.

this screen is an CX3 even through i ordered an CX6 i have no clue what the difference is most likely the stand.

chros

4th July 2021, 14:07

i want later used send bt 2020 option in madVR for HDR -> SDR tone mapping but in native without ever using the remote like every TV before could do.
Then you don't have to do anything special, just setup a profile in madvr with "if (HDR) ... else ..." and set the 2 3dluts. (madvr requires gamma 2.2 3dluts for hdr2sdr tonemapping).
But you don't want to do this on an HDR capable display, but output HDR all the time, but you will see. :)

i heard they support it but i could not yet find it only 22 point.
You need calibration softwares that support this (Calman, ColorSpace, they support the built-in pattern generator) or the previously mentioned aiopylgtv that can generate acceptable 3dluts by LG TV from DisplayCal generated IRIDAS (*.cube) 3dluts (here (https://hub.displaycal.net/forums/topic/lg-c8-lut/page/19/#post-30492) you can read more about it).

BTW. banding in PC mode isn't bad here it doesn't even challenge the xh90 which was perfect in 8 bit mode but it isn't so bad that i should ever notice it with real content. i only tested 8 bit right now i now regret not getting a HDMI 2.1 card just because my last screen 120 HZ was a lie.
Then please test those later that I posted above, thanks.
I don't miss 120Hz, I don't think we need it for movie playback.

chros

4th July 2021, 15:04

I just used the DCI-P3 D65 preset and changed gamma to 2.2 and in the 3D Lut Maker the standard BT2020 profile with gamma 2.2. White Level doesn't matter, I tried 100 nits and 10K nits, no difference.
Rendering intent absolute colorimeric for now but I'll probably try the other ones.
Thanks for posting the images, again. Indeed changing the white level with the DCI-P3 icm doesn't do any change with madvr (while other settings are the same you suggested). But it definitely did when I uploaded the 3dlut to the TV, so who knows why.

Now, about rendering intents (https://displaycal.net/#rendering-intents) (I quickly tried all of them):
- "Absolute colorimetric with white point scaling" (-iaws) would be the correct one technically speaking, but after comparing with the factory 3dlut, it's undersaturated most of (all?) the time, resulting in less punchy image (the other 2 "Absolute" intents are really similar/same)
- - maybe because of WOLED (white subpixel)?
- “Relative” (-ir) looks like Absolute
- “Luminance matched appearance” (-il) looks like Absolute
- “Luminance preserving perceptual” (-ilp) is undersatutes colors and providing weird colors in dark scenes, so it's a no go
- “Perceptual” ones undersatutes colors, so they are no go
- "Preserve Saturation" (-ims) and "Saturation" (-is) work similarly, the latter adds more suturation
- - these look closest to factory 3dlut most of the time (sometimes colors still undersaturated, other times over, but not too distracting)
- - the wierd thing is with these 2 that shadow detail is crushed now: I had to create a gamma 1.0 (!!!) version of these (by setting it both on dic-p3 icm and on 3dlut window) to somewhat match the Absolute ones (and still a bit darker sometimes) :) (how the heck this works ...?)

I will setup profiles for couple of these (-is, -ims, -aws, -ir) in madvr and will switch between them during playback to pick up more issues with them.
Thoughts?

SamuriHL

4th July 2021, 17:19

Well....I'm down and out for quite a while. My 2070 died which was causing the hardware lockups. I'm running on the Intel 630 right now which won't detect that my C8 is an HDR capable display. HAHAHA, kill me now. And of course I have a movie night next weekend with a friend. Guess I'll be using the SHIELD for that one. UGH. Need to hunt down a new video card but I'm not getting a TI and I'm not spending ridiculous money for a 3080. This may take a while.

chros

4th July 2021, 18:31

:D Just get a used 1060 6GB (for less than $100), and you are good to go :)

I have set up the Saturation (-is) gamma 1.0 3dlut as default, images are stunning, after 2 years watching that crap finally we can ejnoy HDR10 again!

j82k

4th July 2021, 19:11

- - the wierd thing is with these 2 that shadow detail is crushed now: I had to create a gamma 1.0 (!!!) version of these (by setting it both on dic-p3 icm and on 3dlut window) to somewhat match the Absolute ones (and still a bit darker sometimes) :) (how the heck this works ...?)

Thoughts?

I would've thought that when both profiles are set to the same gamma you would always get the same result, no matter if it's 1.0 or 2.2 or 2.4, but apparently not? Guess I have to mess with this too.

SamuriHL

4th July 2021, 19:41

I gave my 1060 to my daughter so no getting that back. And I'm not going backwards. I have my laptop with a 2080 so I can always use that temporarily. Just annoying.

chros

4th July 2021, 20:56

I would've thought that when both profiles are set to the same gamma you would always get the same result, no matter if it's 1.0 or 2.2 or 2.4, but apparently not? Guess I have to mess with this too.
Yes, just tried out with "Absolute" and gamma 1.0 , gamma 1.0 is slightly brighter than 2.2

But shouldn't we have to select "BT.2020ST 2084 10000 cd/m2" (from the dropdown) instead of "Bt.2020 color profile" on the 3dlut creation window?
(Although I don't see any difference using "Absolute" rendering intent, after a quick check.)

huhn

4th July 2021, 21:29

Then you don't have to do anything special, just setup a profile in madvr with "if (HDR) ... else ..." and set the 2 3dluts. (madvr requires gamma 2.2 3dluts for hdr2sdr tonemapping).
But you don't want to do this on an HDR capable display, but output HDR all the time, but you will see. :)

3D LUT calibration isn't good enough for me it has to be also calibrated outside of madVR.

for gamut this CX can change the gamut in PC mode but if VRR is triggered it will force auto and i have not yet confirmed if it is wide in this case because the colors are really really off on this device.

if i used 60 HZ 1080 to see how good the scaler is the screen just started to flicker like crazy i'm happy to be not affected by this but i can hardly look at the screen.

PC mode is utterly confusing me with it i can change the gamut setting outside of game preset which is by other TVs needed for "PC mode" but for this TV it doesn't seem to affect input lag and it also doesn't affect 4:4:4 so why is gamut than limited in preset game?

isf expert dark mode is less accurate than game with warm 2.

i'm utterly confused by this device but everything except the 60 flicker issue should be fixable if i take the time the CMS if not buggy seems very powerful.

my first HDR test was very disappointing there is barely any difference between SDR brightness and HDR brightness (with default settings).

the brightness option seem to be the biggest mis engineering i have seen in a while it just rises the black floor extremly on an oled did i get this correctly? what is the point of this...

on the sony TV i couldn't even find a combination of setting where the screen would start clipping black/white levels and the CR will not suffer to much from the settings too this includes brightness 100 and CR at 100.

huhn

5th July 2021, 13:27

10 bit RGB and 8 bit RGB are the same in term of banding is bad but shouldn't be to big of an problem.

btw. i'm starting to loose it:
https://abload.de/img/lgemj8b.png

what the hell are you doing?

oldpainlesskodi

5th July 2021, 13:38

On my LG Oled, I found that the "Smooth Gradation" option on the set actually caused banding lol (full chain)..so now it's off!

huhn

5th July 2021, 13:56

clearly possibile but for testing it needs to be off anyway.

i just found the energy saving thingy turning that off would have given me a much better first impression.

100 nits HDR wasn't impressive at all.

what is the highest none ABL 50% white this TV can do? 260 nits?
the ABL is horrendous as expected. utterly hate stuff like even on FALD i turn the active back light off because it's just bad and this is very similar just in bright scene.

chros

5th July 2021, 15:27

the "Smooth Gradation" option on the set actually caused banding lol (full chain)..so now it's off!
Niiice, never used it, it's switched off here since the beginning. As a matter of fact, almost everything is switched off :)

3D LUT calibration isn't good enough for me it has to be also calibrated outside of madVR.
I see, then for SDR you can use different presets of the TV (Cinema, Filmaker, ISF Bright/Dark), so you don't need the bt2020 flag either.

PC mode is utterly confusing me with it i can change the gamut setting outside of game preset which is by other TVs needed for "PC mode" but for this TV it doesn't seem to affect input lag and it also doesn't affect 4:4:4 so why is gamut than limited in preset game?
Good question, LG always did that.

isf expert dark mode is less accurate than game with warm 2.
That's weird, ISF presets should be the most accurate ones (in normal mode anyway).

the CMS if not buggy seems very powerful.
I wouldn't bother with that for SDR: leave factory 1dlut in place, then create a 3dlut and upload it into different presets.

my first HDR test was very disappointing there is barely any difference between SDR brightness and HDR brightness (with default settings).
LG on older sets defaults Oled light to 80 in SDR which is waaay too high. I use 33 (~120 nits) with gamma 2.4 for night, and 39 (~140 nits) with gamma 2.2 for day if it's necessary. (SDR 3dluts on top of these in madvr.)

the brightness option seem to be the biggest mis engineering i have seen in a while it just rises the black floor extremly on an oled did i get this correctly?
brightness yes, use only oled light, leave brightness/contrast on the preset default (that's their bypass settings).

https://abload.de/img/lgemj8b.png what the hell are you doing?
:D For whatever reason Windows reinit your display all the time (hence madvr sees it as new device).

i just found the energy saving thingy turning that off would have given me a much better first impression. 100 nits HDR wasn't impressive at all.
:D Turn off every bullshit setting, at least for madvr in PC mode.

what is the highest none ABL 50% white this TV can do? 260 nits?
rtings measured (https://www.rtings.com/tv/reviews/lg/cx-oled) more in SDR.
You have to specify the mode you use (and LG can even vary between presets in a given mode! Cinema vs Filmmaker)

the ABL is horrendous as expected.
You can turn off ASBL and GSR in In-Start service menu, these are the ABLs (https://www.avforums.com/threads/lg-oled-c8-owners-and-discussion.2162797/post-28296896) that we know about.

chros

5th July 2021, 15:52

Guess I have to mess with this too.
I have set up the Saturation (-is) gamma 1.0 3dlut as default, images are stunning, after 2 years watching that crap finally we can enjoy HDR10 again!
Unfortunately this sometimes results in oversaturated skin colors, while fire/explosion are still desaturated.

Then I just turned on madvr's saturation feature at +8, and I tried Absolute again: unfortunately it pushes blue too much in dark scenes, so I don't like this either.
But with this (+8) and "Preserve saturation" (-ims) looks promising after couple of clips, I'll use this as default and if I won't notice anything outstanding then I'll never touch this crap again! :)

j82k

5th July 2021, 16:04

Too much saturation in blueish dark scenes is normal for the C8 I think, as I also notice this in SDR and also noticed this when I had the C1 and C8 side by side.
I call it the Blue Plague. :D Still better than the mix of cyan/magenta we had before though...

Perceptual has less saturation in these scenes but it shifts blue too much towards purple.

I found absolute colorimeric HDR to look quite similar when comparing it to HDRtoSDR.

As for gamma, I don't really know. There are minor differences but I couldn't tell which looks more correct.

biship

5th July 2021, 19:58

For the C9 & passthrough, is it best to set 2 calibration profiles, one for HDR & one for SDR and set them as already calibrated and BT.2020 & BT.709 respectively?
Or when using passthrough, is there even any need to set any calibration option and just disable calibration controls for this display?
What about disable GPU gamma ramps, which isn't available when disabling calibration?

chros

6th July 2021, 16:02

@biship: for passthrough it doesn't matter what you set, nothing will be used.

Too much saturation in blueish dark scenes is normal for the C8 I think, as I also notice this in SDR and also noticed this when I had the C1 and C8 side by side. I call it the Blue Plague. :D
Did the C1 also have this?

I'm not sure about sdr here (since I use 3dlut for madvr), DoVi is fine, but HDR10 factory 3dlut is pushing way too much blue. The amount of blue it has is almost like with the "Absolute" 3dluts +8 saturation in madvr! That means by default LG added saturation to the hdr10 3dlut to make it look way punchier.
Take a look at this sample 04-2160p_23fps_hdr0999-l_s02e01_cyan_error.mkv (https://mega.nz/#F!r9N0VK5R!WnnyYsGqwa7M-v1jckw6iQ) (I think I posted this before somewhere, beginning of Lore s02e01), the blue push is almost everywhere. Just checked out the 1080p version of it (madvr + sdr 3dlut) and there's no such blue in it, it's closer the the Saturation intents with hdr10.

As for gamma, I don't really know. There are minor differences but I couldn't tell which looks more correct.
g2.2 looks good for almost all of them, except for the 2 saturation ones, where maybe even g1.0 is not enough (but it brightens other part of the image more then the rest of the rendering intents: this is how the factory 3dlut also work!).

I found absolute colorimeric HDR to look quite similar when comparing it to HDRtoSDR.
It's way undersaturated compared to factory hdr10 3dluts here, see below.

Perceptual has less saturation in these scenes but it shifts blue too much towards purple.
Yes, I don't like this one at all.
But check out "Preserve Saturation" (-ims) g1.0 with +7 saturation in madvr (I lowered down yesterday by 1):
- by default this one also has the slight purple push, but when adding saturation in madvr, it pushes purple towards to blue
- this one look closest how I think images should look like (due to the lack of reference monitor) without having the blue push, rarely pushes more yellow orange than it should
- fire/explosion/sundown look extremely good, no more dried out grass/dirt, etc.

Try the sample above with this, you'll see a drastic difference :)
And I just watched Lights Out (2016) yesterday with it, it looked gorgeous, check it out if you can! (Video has very clean image, lot of rich colors everywhere, perfect hdr representation, and last but not least I actually liked the movie itself :) )

j82k

6th July 2021, 22:10

Did the C1 also have this?

In those problematic blueish scenes, like the lions king, the C1 had less saturation and it was shifted more towards cyan whereas the C8 is shifting towards purple.

I was mostly looking at HDR when comparing them, should've done some SDR too. The C1 is still sitting here boxed up, waiting to get picked up in 2 days. I could unbox it and set it up again but that's way too much effort. :)

I'm not sure about sdr here (since I use 3dlut for madvr)

Funny that you mention this because I just checked my SDR calibration folder and in april/may I did 26 SDR 3DLUTs in displayCal, with all kind of different settings and patch sets, some up to 10K patches. My goal was to fix skin tones + dark blueish scenes. But often the 3DLUT made the dark blues even more saturated or there was a noticeable shift between blue and purple or color banding/posterization. In the end, more dark colored patches would usually mess things up even more.

I guess this could be because of panel drift or because the i1Display Pro being too inaccurate when reading dark colors, or crappy TV processing, I don't know. Some of these 3DLUTs were okayish but I wasn't really happy with any of them. The handmaids tale has lots of blueish dark scenes, this is where I often noticed SDR 3DLUT issues.

Skin tones are easy to fix with a 3DLUT, at least in SDR but getting dark blue colors right is a pain in the ass, at least on my C8.

Take a look at this sample 04-2160p_23fps_hdr0999-l_s02e01_cyan_error.mkv (I think I posted this before somewhere, beginning of Lore s02e01), the blue push is almost everywhere. Just checked out the 1080p version of it (madvr + sdr 3dlut) and there's no such blue in it, it's closer the the Saturation intents with hdr10.

Yeah it's too blue, but watchable. I didn't look at that one with the C1 but I'm sure it would look less saturated and a bit more shifted towards cyan.

Maybe SDR isn't even oversaturated with dark blue scenes and I just have this in my head because of all the messed up 3DLUTs I did. But SDR dark blues are for sure too much shifted towards purple on my C8.

It's way undersaturated compared to factory hdr10 3dluts here, see below.

I mostly did comparisons with darkish scenes, so what you say could very well be the case in brighter scenes or with colored highlights.

Yes, I don't like this one at all.
But check out "Preserve Saturation" (-ims) g1.0 with +7 saturation in madvr (I lowered down yesterday by 1):
- by default this one also has the slight purple push, but when adding saturation in madvr, it pushes purple towards to blue
- this one look closest how I think images should look like (due to the lack of reference monitor) without having the blue push, rarely pushes more yellow orange than it should
- fire/explosion/sundown look extremely good, no more dried out grass/dirt, etc.

Try the sample above with this, you'll see a drastic difference
And I just watched Lights Out (2016) yesterday with it, it looked gorgeous, check it out if you can! (Video has very clean image, lot of rich colors everywhere, perfect hdr representation, and last but not least I actually liked the movie itself )

Still too purple for me, I don't know how much panel variance there is when it comes to hue/saturation but on my C8 for skin tones and brownish-orange colors to look correct I need to set the tint slider towards green.
I already knew that before, but having had the C1 next to the C8 just confirmed this as skin tones looked similar with tint at about G8-G10. I also did this comparison when I had the GZ2000 and back then the C8 also required a tint adjustment towards green to look comparable, don't remember how much though.

Now the huge drawback when setting tint towards green is that dark blueish scenes now look worse because they'd need to have the tint slider set towards red.

So it's either 'dark blueish scenes looking purple' or 'skin tones looking too red', or some of each with tint at default 0.

I tried to fix this with the CMS controls but they do almost nothing to dark colors and barely anything to skin tones even when set to max.

I think I'm done with testing and 3DLUTs for now, at least until something promising emerges... :)

The C8 issues aren't too bad and at least my C8 has no near-black overshoot with black levels set to faintly glow. This tweak doesn't work on the C1 because of the heavy dithering and because it doesn't use the white sub-pixel below 3% or so.

chros

8th July 2021, 20:12

Funny that you mention this because I just checked my SDR calibration folder and in april/may I did 26 SDR 3DLUTs in displayCal, with all kind of different settings and patch sets, some up to 10K patches.
...
I guess this could be because of panel drift or because the i1Display Pro being too inaccurate when reading dark colors, or crappy TV processing, I don't know.
:D I can confirm this, I also tried bunch of profiling sessions, I can get a good one let's say 1 out of 10 :D Also ABL can be an issue (at least on mine, GRS). How the hack this works ... :)
I managed to make a good one back in the day, so that's what I use, probably until the TV dies :)

I mostly did comparisons with darkish scenes, so what you say could very well be the case in brighter scenes or with colored highlights.
It's not even with highlights, just normal scene, like sand/grass/tree/etc.

Still too purple for me, I don't know how much panel variance there is when it comes to hue/saturation
It can be, or you know, the used 1dlut can also modify things.

but on my C8 for skin tones and brownish-orange colors to look correct I need to set the tint slider towards green.
I already knew that before, but having had the C1 next to the C8 just confirmed this as skin tones looked similar with tint at about G8-G10. I also did this comparison when I had the GZ2000 and back then the C8 also required a tint adjustment towards green to look comparable, don't remember how much though.
Interesting. Just as you said, it's hard to create 1 that is good for everything :)

I think I'm done with testing and 3DLUTs for now, at least until something promising emerges... :)
Same here as well (probably :)), but with this it's lightyears ahead how it was before! :D

The C8 issues aren't too bad and at least my C8 has no near-black overshoot with black levels set to faintly glow. This tweak doesn't work on the C1 because of the heavy dithering and because it doesn't use the white sub-pixel below 3% or so.
Interesting, thanks for the info. As I mentioned couple of time, the macroblocking doesn't bother me at all if it's present (probably it's not that disturbing here).

1 last thing about 1dluts: do you remember when I uploaded hdr10 1dluts into DoVi presets and said that couple of them destroyed the "neon" color look (I don't know how to describe it otherwise, it's extremely visible when you switch between presets), resulting in pale colors.
Now, since we finally have a clean image with bloody hdr10 as well, I uploaded couple of those 1dluts, and I noticed the same thing!
What's your take on this, what can be the cause? Can it be the misaligned Gray scale balance?

ashlar42

9th July 2021, 13:32

Could somebody be kind enough to do a "differences between 1dlut and 3dlut for dummies" for me (the dummy)?

quietvoid

9th July 2021, 13:52

I'm not an expert but essentially 1D LUT is for greyscale and whitebalance (brightness adjustments, whitepoint), while 3D LUT is for aligning colors to what they should actually look like.

SamuriHL

9th July 2021, 20:03

That sums it up. If you want a deeper look at it:

https://www.lightspace.lightillusion.com/luts.html

That should help.

ashlar42

9th July 2021, 21:11

Thanks to both!

chros

10th July 2021, 11:42

LG on older sets defaults Oled light to 80 in SDR which is waaay too high. I use 33 (~120 nits) with gamma 2.4 for night, and 39 (~140 nits) with gamma 2.2 for day if it's necessary. (SDR 3dluts on top of these in madvr.)

I think I'm done with testing and 3DLUTs for now, at least until something promising emerges... :)
Then try this one out! :) This is about SDR.

Since madshi said, he won't deal with SDR2HDR for now and we have 0 black (meaning we can raise Oled light to the stars without elevating black floor), I was curious how HLG works (that's the closest to SDR+HDR), that's how I found this doc (https://www.avsforum.com/threads/improving-madvr-hdr-to-sdr-mapping-for-projector.2954506/post-60869970).

If I understand it right, they claim (by visual testing), that the knee point (they chose 80% input signal, ~58nits @100nits White and g2.4) is not that important than the actual steepness above it, and they try to map diffuse white (90%/95% input signal, ~78/88 nits @100nits White and g2.4) to 203 nits and the rest is highlight.
The interesting part is the graph on page 20, they almost follow the same curve until the knee point (they go a bit higher by mapping the 58nits to ~70nits), from there they use different curve until they reach diffuse white, then another one above that.

So, I was curious what options we have to accomplish something like that with our sdr 3dluts. As it turned out not much :)
- we have 0 black, and although faking black/relative/absolute/offset params do work on Calibration tab of DisplayCal (argyll cms), but even with the lowest settings the raised black is getting annoying

That leaves us with standard gamma setting:
- I made this sheet (https://docs.google.com/spreadsheets/d/1-pUyOz1idTL-KhMhc1Kk5yS7Ir4bgWZl2ABCNOHFqis/edit?usp=sharing) to see what brightness we get for different input signals with different gamma and peak settings (we want to compare it with the 100/120 nits values)
- want to use as high (darker) gamma as we can without black crush, because higher luminance range will be brighter anyway
- - the green lines are the gamma that I picked for each peak, by trying to match the 0,05 value to the 100 nit one (didn't get black crush with them after checking it with various scenes)
- if we use really high Oled light (peak) (>271 nits) then even the suggested minimum value of diffuse white (90%) will be really bright (>203 nits), so it's unnecessary

So, I experimented a bit:
- at first I picked the 208 nits peak one (Oled light 70 here) and I used it for couple of months as proof of concept
- later I went for 271 nits (98 Oled light here) meaning 203 nits ~ 90% Input signal, but we drive our panel really hard this way and red/orange color was too saturated sometimes (maybe it was the cause of the generated 3dlut)
- finally I settled down with 234 nits (85 Oled light here) meaning 203 nits ~ 95% Input signal

So with 234 nits (85 Oled light here):
- I created a g2.68 3dlut (and a g2.48 for bright daylight): set Maximise RBG luminance at patch generation!
- it has more than double of the peak of the standard sdr (but due to gamma the increase is way less with lower input signals)
- no balck crush
- 80% input signal (knee) is around 129 nits with this gamma
- no banding/posterization/other anomaly
- no color desaturation (even more saturation)
- brighter scenes are not disturbing during the night either

Due to the increased luminance range, we get higher sharpness, punchier / more lifelike images, while dark scenes don't suffer. Where this really shines (obviously) is with high contrast scenes including lights/shiny objects/reflections.

So, I have set up Cinema SDR preset with Oled light 85 and created a profile in madvr for this g2.68 3dlut.

(I also tried with my original 120 nits (33 Oled light here) profile:
- creating HLG curve from the same profile: sometimes madvr didn't accept the 3dlut, other times it did, bit the result wasn't impressing, I guess due to the peak 120 nits in the 3dlut
)

Note about burnin: these guys made ADL measurements with sdr movies (http://projectiondream.com/en/movie-brightness-adl-contrast-measurements/) (just as we did with HDR10 with madmeasure), and the brightest one (snowboarding doc :) ) had 32% ADL, while the average of the rest is ~10% ADL.
So, although we drive our panel harder with this (e.g. compared to 120 nits peak), but this won't be an issue with standard (low APL) video content (e.g. 50% input signal of SDR with the latest settings above is ~37 nits, while the same with HDR10 at 700 nits peak is ~100 nits).

oldpainlesskodi

10th July 2021, 17:05

chros

11th July 2021, 12:28

)I also tried:
- creating HLG curve from the same profile: sometimes madvr didn't accept the 3dlut, other times it did, bit the result wasn't impressing, I guess due to the peak 120nits in the 3dlut
)
...
Next I'll try out some synthetic 3dlut with HLG curve, maybe ... :)
I did, in short I'll use the settings that I came up with previously (70 oled light with g2.64 3dlut) until madvr has sdr2hdr10 feature:
- synthetic 3dluts doesn't have any colorimetric data, colors are slightly desaturated, it has less impact
- - it was a good test for HLG, the peak brightness has a big impact of the final image (e.g. 250 vs 1000)
- same with the 3dlut I use now with HLG: it only has 120 nits peak, so maybe that's why
- we can't profile this crap at 90/100 oled light due to instability, even then maybe I wouldn't prefer the HLG 3dlut variant

So, after a while of resisting PC mode on the LG because it removed some options that I thought I found useful, I decided, as an experiment to switch to PC mode and make some adjustments, and now I cant go back to 4.2.0 vs 4.4.4...please help me un-see the the light!
So, do you like PC mode with chroma 4:4:4? :) (it's 4:2:2 vs 4:4:4, not 4:2:0) If so, tell @j82k :)
What options do you miss with PC mode?

@huhn, how is it going with your new toy? :)

huhn

11th July 2021, 15:41

just finished the first test calibration:
3D LUT: https://drive.google.com/file/d/1GT0hyQbN0u_M7C0GkYWgQ_b46kXUWc7W/view?usp=sharing
22 point ""calibration: https://drive.google.com/file/d/1yq4L-IbGrH76YkNJ1E4TgughskwpIRIi/view?usp=sharing

not my best work but this was a test the manual calibration took me while i did never verified it "only" fixing white point for the gray scale.

i will not play around with the internal 3DLUT yet. i have no real banding issue or close to black blocking thingys lucky bad tests?

but my screen/AMD/windows11 do something very strange 2-6 pixel are cut of the left side and the right side looks like it has space for these pixel...
edit: it fixed it self after a couple of resolution changes

oldpainlesskodi

11th July 2021, 19:13

@ chros - thanks for the correction lol.

Yes, I prefer 4.4.4(:!), but it's a shame that all advanced controls locked unless in ISF mode in SDR,.... in HDR mode advanced controls locked out completely regardless of mode (color gamut set to wide, peak brightness etc etc)...

j82k

12th July 2021, 23:28

...lots of of words ....

Sorry for not replying earlier. I'm very busy with other things so I'm currently not messing with TV stuff at all.

The C1 is gone and I'm getting a full refund. To be honest I think I would've regretted this TV forever had I kept it. :devil: It's better than the C8 in some regards but worse in others, so I'm happier with not paying a lot of money for a TV that is only a compromise. :)

I wish I wouldn't care about near-black overshoot, because then I would just buy a panasonic oled and be happy but I hate that crap and every time I see it, in my mind I bought an expensive faulty piece of crap TV.:eek: I just can not stand it and I hope I will not buy another oled until this is 100% solved on a hardware level.

About dolby vision:
I have no way of measuring DV and I rarely watch DV. The only thing I did some time ago was measure a 50% DV patch that I found and with the same 1DLUT that I use for HDR it was way too bright... So to me it seems DV needs a different 1DLUT?

Asmodian

13th July 2021, 08:30

Are you looking for HDR settings? I just use passthrough on my 48CX. :o

huhn

13th July 2021, 11:43

you are not disappointed on these screens HDR passthrough?

did you try to use SDR -Y HDR tone mapping to improve the situation?

quietvoid

13th July 2021, 13:35

For 95% of the (current) content there will be little to no difference between passthrough and tonemapping with madVR.

nevcairiel

13th July 2021, 15:12

I got a 65G1 today, going to need to break out the meter soon and figure out how all this worked, been a while since i had to touch this for my old trusty 1080p TV.
Won't have it properly mounted on the wall until next week, and going to need to install a 30-series card in the HTPC (which I already have, too) to make full of use of all the HDMI (and the 1060 i had previously is certainly at a limit upscaling to 4K), hopefully i can mess around with some measuring on the weekend.

chros

13th July 2021, 16:10

Congrats, Nev, ask any question ... :)

Asmodian

13th July 2021, 19:28

you are not disappointed on these screens HDR passthrough?

I don't watch much HDR content in the first place, but in all my testing of various tone mapping settings I found some scenes look better but others look worse. Also, it would look a bit better than passthrough but sometimes it was much worse than passthrough, the upside was smaller than the downside.

I haven't been keeping up with all the beta builds, so maybe there are some settings that are noticeably better across all content. :p

did you try to use SDR -Y HDR tone mapping to improve the situation?

SDR -Y HDR?

huhn

13th July 2021, 22:56

I don't watch much HDR content in the first place, but in all my testing of various tone mapping settings I found some scenes look better but others look worse. Also, it would look a bit better than passthrough but sometimes it was much worse than passthrough, the upside was smaller than the downside.
good to know not a an of the current HDR -> SDR situation too.
I haven't been keeping up with all the beta builds, so maybe there are some settings that are noticeably better across all content. :p

i don't really follow them to is more about working around the ABL as much as possible

SDR -Y HDR?

german keyboard and me failing spectacular

chros

14th July 2021, 09:56

just finished the first test calibration
...
i will not play around with the internal 3DLUT yet. i have no real banding issue or close to black blocking thingys lucky bad tests?
...
but my screen/AMD/windows11 do something very strange 2-6 pixel are cut of the left side and the right side looks like it has space for these pixel...
edit: it fixed it self after a couple of resolution changes
Did you turned off Pixel shift in the TV? (That can do things like this)
I'm not bothered by macroblocking either, it depends on lot of things.
Interesting that you calibrated it at 200 nits. How much did you measure in Cinema preset HDR10 mode with your CX?

anyone have a 48cx setting they can share? like upload the settings.bin or registry , im lost on how to properly set this, currently have a new rig with a 5950x and 3090 wanted to try if thier anyone with a good max settings to try thanks
Take a look at the link in my signature for settings for 1060 6GB. You can crank them up if you like, but I doubt you will see any difference (only on your electricity bill :) ).

I found some scenes look better but others look worse. Also, it would look a bit better than passthrough but sometimes it was much worse than passthrough, the upside was smaller than the downside.
Yes, that's why I disabled DTN in madvr for my default profile half a year ago :)

Sorry for not replying earlier. ... The C1 is gone ...
No worries, me too. I agree, what's the point to pay lot of money for another device with full of issues, we already have one :D

I wish I wouldn't care about near-black overshoot, because then I would just buy a panasonic oled and be happy but I hate that crap and every time I see it, in my mind I bought an expensive faulty piece of crap TV.:eek: I just can not stand it and I hope I will not buy another oled until this is 100% solved on a hardware level.
I doubt it will be ever solved due to the inherent nature of the problem is the technology itself, WOLED :)
Panasonic is not acceptable for me either, unless they fix(ed ?) judder free 23p with chroma 4:4:4, that I doubt ... Let alone, the tools/ecosystem around LG is so well built at this point ...

About dolby vision:
I have no way of measuring DV and I rarely watch DV. The only thing I did some time ago was measure a 50% DV patch that I found and with the same 1DLUT that I use for HDR it was way too bright... So to me it seems DV needs a different 1DLUT?
Can you send me over that patch via pm? Thanks
It depends on the preset you tired: which one was it? (game preset is 150 nits lower here on B8)
I can't tell it now, because the 1dlut I use (Can you send me over that patch via pm? Thanks) produce brighter normal scenes, probably due to the elevated gamma from 30% (which I had to do to don't have fog/veil with factory 3dlut in normal/bright scenes).
So I guess, I have to go back to 1dlut creation and try out flat gamma 2.2 again with our synthetic hdr10 3dlut and DoVi. But it's just so finicky to deal with it ... :)

huhn

14th July 2021, 11:42

Did you turned off Pixel shift in the TV? (That can do things like this)
off
I'm not bothered by macroblocking either, it depends on lot of things.
i tested for worse makro blocking not makro blocking if ther eis makro blocking i can't blame the TV. it the TV would do that i would be very annoyed.
Interesting that you calibrated it at 200 nits. How much did you measure in Cinema preset HDR10 mode with your CX?
i did not measure HDR10 cinema because the problem is not the brightness is more than 200 sometimes much more it's the inconsistence that's the problem.

Asmodian

14th July 2021, 19:01

so i was hoping to just copy from someone, i remember it can be done by registry or a setting.bin

The third post of the link in my signature does have settings.bin for v0.92.17.

huhn

16th July 2021, 12:54

nice test: https://www.youtube.com/watch?v=Qtve0u3GJ9Y

good to see that i'm not alone on the out of the box accuracy.
i disagree with HDR which is a joke in my opinion.

and the screen is sometimes still cut of toggling the screen shift fixes it but still bad design maybe i need to update the firmware...

quietvoid

16th July 2021, 14:09

and the screen is sometimes still cut of toggling the screen shift fixes it but still bad design maybe i need to update the firmware...

Do you have Just Scan enabled?
Might be different for new models, it's an aspect ratio setting.

Not sure how the screen could get cut if shifting is disabled otherwise.

huhn

16th July 2021, 18:01

i would know if it would over scan no it'S the pixel shift been buggy toggling it on off fixes the issue.

Klaus1189

16th July 2021, 18:04

huhn

16th July 2021, 19:24

the C1 is supoposed to have a new panel all measurements i have seem show the c9, cx, c1 are pretty much the same.
there are differents but they are very low
they are the"flag ships" models from 2019, 2020 and 2021

Asmodian

16th July 2021, 19:36

i would know if it would over scan no it'S the pixel shift been buggy toggling it on off fixes the issue.

If you disable Quick Start+ it fixes the pixel shift issues, at least on my CX. :p

huhn

17th July 2021, 00:11

i give it a try.

Magik Mark

17th July 2021, 04:10

quietvoid

Any Feedback from the latest madvr build? Is it even worth it to upgrade?

quietvoid

17th July 2021, 14:55

quietvoid

Any Feedback from the latest madvr build? Is it even worth it to upgrade?

Not really any difference currently, might be worth it when we start getting improvements on the flickering (or rather when we know what values are good)

freebits

18th July 2021, 07:52

Hi,

This is my 1st posting since joined 2007. lol.

One thing I am interested in is applying madVR 3d lut to LG OLED in PC HDR mode and wonder if it's ok if I create 3DLUT SDR DCI-P3 via displaycal or Calman and apply it to madVR for HDR? (I have Calman and K10 meter btw.)

I've tested SDR DCI-P3 3D LUT in 4K projector via madVR pixel shader and it worked well. But LG OLED SDR mode does not boost up to 700 nits giving only max 400 nits so HDR-to-SDR via pixel shader does not give higher impact.

What solution do you advise to get a calibrated HDR in PC? no 3d lut solution for madVR HDR and just HDR calibration via Calman?

Thank you

chros

20th July 2021, 09:32

guys question, when i play a movie thats in hdr the color is washout when not in fullscreen mode, its only ok if in fullscreen, anyway around this?
Recent nvidia drivers work like this, which driver and card do you have?

One thing I am interested in is applying madVR 3d lut to LG OLED in PC HDR mode and wonder if it's ok if I create 3DLUT SDR DCI-P3 via displaycal or Calman and apply it to madVR for HDR? (I have Calman and K10 meter btw.)

I've tested SDR DCI-P3 3D LUT in 4K projector via madVR pixel shader and it worked well. But LG OLED SDR mode does not boost up to 700 nits giving only max 400 nits so HDR-to-SDR via pixel shader does not give higher impact.

What solution do you advise to get a calibrated HDR in PC? no 3d lut solution for madVR HDR and just HDR calibration via Calman?

Which TV do you have?
We had problems with 2018 LG TVs, that's why we tried/use synthetic 3dlut with hdr output.
But if you don't have any issues with the factory 3dlut processing of your TV then just leave it as it is, or can calibrate the TV with calman.

freebits

20th July 2021, 13:56

Recent nvidia drivers work like this, which driver and card do you have?

Which TV do you have?
We had problems with 2018 LG TVs, that's why we tried/use synthetic 3dlut with hdr output.
But if you don't have any issues with the factory 3dlut processing of your TV then just leave it as it is, or can calibrate the TV with calman.

I have C9 and CX.
I'm fine with HDR 3d lut for my 4k players via Calman matrix lut but for GPU & madVR, I don't know how to create 3D LUT for PC HDR mode.
I can do SDR BT2020 lut via madVR's HDR-to SDR conversion but it won't work effectively for LG OLED TV since SDR mode is not brighter than HDR, right?
neither HDR BT2020/P3 lut is possible since measuring madVR's HDR is unstable, no? Is it successful for you when you do HDR 3D LUT via madtpg?

SamuriHL

21st July 2021, 19:53

Well nVidia RMA'd my 2070. We'll see how that goes. I've not been able to secure a 30xx card so hopefully I can at least get myself back to where I was and get testing again.

chros

22nd July 2021, 20:06

I can do SDR BT2020 lut via madVR's HDR-to SDR conversion but it won't work effectively for LG OLED TV since SDR mode is not brighter than HDR, right?
That's correct.

I have C9 and CX.
I'm fine with HDR 3d lut for my 4k players via Calman matrix lut but for GPU & madVR, I don't know how to create 3D LUT for PC HDR mode.
...
neither HDR BT2020/P3 lut is possible since measuring madVR's HDR is unstable, no? Is it successful for you when you do HDR 3D LUT via madtpg?
You can try this (https://www.avsforum.com/threads/madvr-argyllcms.1471169/post-60847806), but there's an issue in this case as the post below it suggests, or you can try to create one in calibration mode.
In short, I didn't like the result, so just ended up using synthetic 3dlut: you can read from here (https://forum.doom9.org/showthread.php?p=1946716#post1946716) (2 pages)
So, if you are satisfied with 3dlut of the TV then no need to create one for madvr, unless you want to experiment with it.

And that's how I watch SDR (https://forum.doom9.org/showthread.php?p=1947198#post1947198) :)

freebits

24th July 2021, 16:57

That's correct.

You can try this (https://www.avsforum.com/threads/madvr-argyllcms.1471169/post-60847806), but there's an issue in this case as the post below it suggests, or you can try to create one in calibration mode.
In short, I didn't like the result, so just ended up using synthetic 3dlut: you can read from here (https://forum.doom9.org/showthread.php?p=1946716#post1946716) (2 pages)
So, if you are satisfied with 3dlut of the TV then no need to create one for madvr, unless you want to experiment with it.

And that's how I watch SDR (https://forum.doom9.org/showthread.php?p=1947198#post1947198) :)

Thank you for your detailed explanation. I'll check each link and comment about them later.

razierklinge

31st July 2021, 00:19

Hey chros, I haven't been on this thread for a long time, but I was one of the people complaining about really bad cyan/purple color bands in HDR last year.

I skimmed through the new pages in the thread and it seems like you might have found a fix for this problem (yay!). But there were over 50 new pages of comments so it's a lot to go through. Is there any specific comment that has the instructions for implementing your fix? Would love to be able to watch my local HDR content again.

chros

31st July 2021, 09:39

Just read my quoted post above yours, the link is there and follow the 2 pages discussion. Which TV do you have?

ashlar42

31st July 2021, 11:53

Ok, I need help.....I'm BLIND!!! lol.

So, after a while of resisting PC mode on the LG because it removed some options that I thought I found useful, I decided, as an experiment to switch to PC mode and make some adjustments, and now I cant go back to 4.2.0 vs 4.4.4...please help me un-see the the light!

Serious note - why hide 4.4.4 that makes it so hard to enable on these panels at a cost of losing other options?
I always forgot to ask about this... I see you talk of PC mode as the way to unlock 4.4.4. And I don't get it. While not in PC mode, if I set RGB full the "mash green button" info panel states that I'm in RGB mode. What you write means it's not full chroma resolution that way?

oldpainlesskodi

31st July 2021, 12:27

Yeah, only PC mode displays 4.4.4 (check here and download the image and open in paint3d to test - https://forum.doom9.org/showpost.php?p=1673852&postcount=25082 )

ashlar42

31st July 2021, 16:33

Yeah, only PC mode displays 4.4.4 (check here and download the image and open in paint3d to test - https://forum.doom9.org/showpost.php?p=1673852&postcount=25082 )Ok, I have checked and you're right.

Does this matter for video playback? Isn't all video content available encoded at 4:2:0? Is gaming your worry or is madVR affected somehow?

oldpainlesskodi

31st July 2021, 16:58

ashlar42

31st July 2021, 18:34

I don't game, my rig is mainly a HTPC. Does it matter for video playback via MadVR? - that's down to the eye of the beholder I guess, but MadVR does an upscale and conversion from 4.2.0 to 4.4.4.Ok, so it's a matter of adding stuff to the source (which might be eye pleasing not saying the opposite) more than risking to lose info.

For me, I notice a difference, even at normal viewing distance - even did a blind test with the wife, and she picked up on it. But, YMMV.
With normal content or patterns?

oldpainlesskodi

31st July 2021, 18:37

TV shows and Movies, but like I said, let YOUR eyes be the judge.

ashlar42

31st July 2021, 19:00

TV shows and Movies, but like I said, let YOUR eyes be the judge.
Sure. I was just curious. Hats off on the above, I'm pretty sure I would be unable to tell the difference.

I'm much more sensitive to movement imperfections.

QBhd

31st July 2021, 19:24

LOL... we all have our own deal breakers... I personally don't like the jitter of OLED and as such I use TruMotion with a 2/0 setting. Another for me is banding in HDR, so I do NOT use PC mode. I hate not having 4:4:4, but the other two are literally deal breakers for me.

QB

huhn

31st July 2021, 20:40

these screen should be burned for the terrible coating.
even the much better CR on these display levees me cold compared to a static 5 K CR display is it better no question but whatever.

the test ufo is just an work of art on these displays while it is not the sharpest it has zero over drive or other issues it's absolutely clean i may even say better than CRT.
the screen uniformity is is also unmatched.

chros

31st July 2021, 20:51

Anyone with newer models (2020/2021) feel free to test these (can be done in 1 hour), thanks.

A. Prereq:
- for near black testing scenarios use pitch black room and let the display warm up for at least 10-20 minutes beforehand
- presets on TV: use Filmmaker/HDR10 Cinema (not home) for HDR10, and Cinema (not home) for DoVi
- - reset the used preset, disable all unnecessary processing, leave brightness/contrast/oled light at their preset defaults, set D65 color temp (if it's not set) and adjust black level to your settings (low/high)
- for HDR10 use PC mode (chroma 4:4:4)
- - during testing pause playback, and switch between normal and PC mode to see differences
- PC output for HDR10: RGB Full 10/12bit output in nvidia panel + madvr set to 10bit or higher (@24/25/120Hz)
- - use passthrough in madvr for HDR10
- with DoVi test: you can switch between TV-led or player-led of you have an Oppo (needs a restart, more info (https://www.avforums.com/threads/the-ultimate-oppo-media-device.2207229/post-28293127))
- - you can also use Plex app and plex server on PC to play DoVi files back on the TV, or via USB, or any capable external devices (but all of these can behave differently!)
- - I don't have chroma 4:4:4 with DoVi with my B8, not even in player-led via Oppo, but it triggers the usual PC-mode issues here!

B. HDR10 in PC mode (chroma 4:4:4)
1. what about banding in HDR10 PC mode? test it with Mehanik's HDR10 test patterns (https://www.avsforum.com/threads/hdr10-test-patterns-set.2943380/) ("03. Grayscale\02. Ramps\01. Diagonal Ramp 400nit.mp4")
2. near black posterization in PC mode? (last part of this post (https://forum.doom9.org/showthread.php?p=1883108#post1883108) for sample and description)
3. black clipping in PC mode? (use "01. Black level\01. black-level-v1.mp4" from Mehanik's HDR10 test patterns)
4. bt2020 color processing issues? (more info (https://forum.doom9.org/showthread.php?p=1925528#post1925528), you need Netflix for this test to be able compare it.)

C. DoVi
1. does DoVi have raised black and still crush near black at the same time? use Stacy's profile8 single track version (https://forum.blu-ray.com/showpost.php?p=18984135&postcount=928) (original profile7 dual track version (https://www.avsforum.com/threads/spears-munsil-uhd-hdr-benchmark-disc-discussion.3075780/page-43#post-59634582)), (or alternatively DoVi samples from here (https://forum.doom9.org/showthread.php?p=1931241#post1931241)), more info (https://forum.doom9.org/showthread.php?p=1932220#post1932220)
- look for elevated black level (there's no true black on 2018 models with DoVi!)
- look for "glowing" black: if that's the case then the underlying 1dlut is responsible for this

D. A possible answer:
- B1: "Not much difference between normal and PC mode, PC mode shows a bit more banding"
- B2: "I haven't noticed much issues, except for @03:25 part that has posterization in PC mode"
- B3: "all blinking above 64 in PC mode"
- B4: "I don't have Netflix" or "I don't see any difference" or "hdr10 looks green/way worse"
- C1: "Using USB + internal player: there's no true black (the whole thing is elevated) with this pattern, >=72 are blinking, plus there's "glowing" in the middle of the screen"

ashlar42

31st July 2021, 20:58

Another for me is banding in HDR, so I do NOT use PC mode. I hate not having 4:4:4, but the other two are literally deal breakers for me.

B. HDR10 in PC mode (chroma 4:4:4)[/U][/B]
1. banding should be largely reduced in HDR10 PC mode, test it with Mehanik's HDR10 test patterns (https://www.avsforum.com/threads/hdr10-test-patterns-set.2943380/) (03. Grayscale\02. Ramps\01. Diagonal Ramp 400nit.mp4)

Since the above two quotes seem stating opposite things... is this because what QBhd states applies to older panels and there's hope things got better with newer panels?

chros

31st July 2021, 21:01

Since the above two quotes seem stating opposite things... is this because what QBhd states applies to older panels and there's hope things got better with newer panels?

That's correct, I copied it from an older post, sorry :) Yes, I/we have 2018 models, that's what I compared to. (Just edited the post.)

chros

31st July 2021, 21:59

Added B.3. small test, on our 2018 models only >=76 are visible.

oldpainlesskodi

1st August 2021, 08:31

@ Chros:

Might just be my eyes, but I see none of the issues described for b1,2,3, well, not that jumps out at me anyway.

I can't comment on C as I don't have any DV content.

MadVR straight passthrough set at 10bit, and full range chain, with the GPU set at 10bit, panel in PC mode.

chros

1st August 2021, 12:00

Thanks, but this is not useful :) Have you used the mentioned samples? Because you have to! :) Do it in a pitch black room and switch between normal and PC mode with hdr10.
About DoVi: yes, you do have at least one :) Copy it to a pendrive and play it back with the internal player of the TV if you don't have any other device. (I added more description for DoVi test.)

A possible acceptable answer could be:
B1: "Not much difference between normal and PC mode, PC mode shows a bit more banding"
B2: "I haven't noticed much issues, except for @03:25 part that has posterization in PC mode"
B3: "all blinking above 64 in PC mode"
C1: "Using USB + internal player: there's no true black (the whole thing is elevated) with this pattern, >=72 are blinking, plus there's "glowing" in the middle of the screen"

oldpainlesskodi

1st August 2021, 13:17

Ok, will check again and come back, at some stage.

And, yes, I used all the samples you provided/pointed to.

Huhn - I feel your pain..

huhn

1st August 2021, 18:14

if it could be downloaded in a sense able time i would be done already i mean this mega link takes hours to prepare a download...
i could upload this faster.

nevcairiel

1st August 2021, 21:43

The mega link for B.2? Thats downloading quite reasonably for me. I'll try to run the tests the next few days. Although my room is never super pitch black, but i'll try to remember to do it at night :D

huhn

1st August 2021, 23:02

the mega things prepares at 600 kbit i have no clue how fast it will be when it is done. my internet has no influence on this.

QBhd

2nd August 2021, 22:04

Wow... 5200+ hours on my C8... I watch far too much TV :P

QB

chros

3rd August 2021, 11:29

Added B.4. test for color processing issue (hopefully the last one). This is also a mega link (but not that big) :) If you downloaded the required samples, and you have better alternative server (that stores the files for years) then link it and I'll add to those posts as backup links.

Although my room is never super pitch black, but i'll try to remember to do it at night :D
Yep, you have to do B2 and C1 in a pitch black room (during the night, blinds down, no lights), otherwise you won't see crap, those are so low light samples :)

Wow... 5200+ hours on my C8... I watch far too much TV :P
I will check mine, but I think it's around 7000 :) The usage hours are ~doubled since the sh1t hit the fan more than a year ago ...

ashlar42

3rd August 2021, 15:06

So, I have calibrated my 77C1 with ColourSpace. At video levels (16-255), as suggested. I have madVR at limited, GPU full, TV limited.

When I pause a video, it’s like madVR switches to full levels (0-255). When I resume playback, levels return to normal (limited) in around a second (I haven’t timed it).

Is this normal behaviour?

chros

3rd August 2021, 15:09

No, not for the old driver behaviour. Is this with hdr10 or sdr?

ashlar42

3rd August 2021, 15:12

No, not for the old driver behaviour. Is this with hdr10 or sdr?
SDR. What do you mean by “old driver behaviour“?

chros

3rd August 2021, 18:46

I thought you meant hdr10, where nvidia changed the way it works in windowed/fullscreen. So I don't think it's right with sdr, but other users have to check who has the new driver.
But since your GPU is at full, maybe it's a madvr fault? which version do you use?
I can try it out later/tomorrow once more with sdr. (I use full chain with sdr, but limited/full/limited with hdr10 and I don't have this problem.)

ashlar42

3rd August 2021, 23:17

It’s Kodi DS Player, for whatever reason. In MPC-HC it doesn’t happen.

chros

4th August 2021, 07:19

Oh, I see, I don't use kodi, maybe someone else.

huhn

6th August 2021, 00:37

- - reset the used preset, disable all unnecessary processing, leave brightness/contrast/oled light at their preset defaults, set D65 color temp (if it's not set) and adjust black level to your settings (low/high)
- for HDR10 use PC mode (chroma 4:4:4)

what's the point of this why do i need to use a setting with avg DE of 10?

there a 4 files:
- banding: e.g. red light of torches around @03:20
- added noise: @00:32, around @01:16
- posterisation: clouds in dark scenes, @03:20, @03:25, etc.
which file at what time?
i mean the banding file is mostlikely for the banding test and there is absolutely no difference between 3.20 on PC mode or not it's looking fake on both but on the other hand there are no torches so...

chros

7th August 2021, 12:43

what's the point of this why do i need to use a setting with avg DE of 10?
Because the whole HDR10 part of the test is about PC mode.

there a 4 files: which file at what time?
Good point, this one: 01-2160p_23fps_hdr0686-thgm_banding.mkv (I edited the old post as well). Times are specified in the old post (https://forum.doom9.org/showthread.php?p=1883108#post1883108).

i mean the banding file is mostlikely for the banding test and there is absolutely no difference between 3.20 on PC mode or not it's looking fake on both but on the other hand there are no torches so...
Torches are the red lights attached to the guns. Check out e.g. @03:25 when soldiers just arrived at the bottom of the building, wall in the background, there's a statue: here the background looks like an oil painting in PC mode :)

huhn

10th August 2021, 15:27

sorry i don't get the expected result.
it always looks fake not like banding but red is missing compared to tone mapping. the statue is at 3.32 at least on my PC.
is a generally noisy image that misses the red light effect.

HDR on my OLED looks just generally worse than SDR even with stock tone mapping something i'm not used too and i only know for comparsion a dim 600 nit 100% white LCD with 5.7 K CR.

my current assumption is that i get the bad result outside of PC mode too but i can't say that the result is banding...

flaviowolff

26th August 2021, 20:34

Has anyone been able to compare the new trumotion "cinematic movement" preset with Madvr's smooth motion?

ashlar42

28th August 2021, 16:32

Has anyone been able to compare the new trumotion "cinematic movement" preset with Madvr's smooth motion?
If you have specific source material for me to check, I'd gladly do a test for you.

chros

8th September 2021, 19:32

There's a new method (https://www.avforums.com/threads/lg-oled-c8-owners-and-discussion.2162797/post-29598721) to downgrade LG firmware even on newer (v5.x.x? ) firmwares in which LG disabled that possibility for the average user. Works on 2018 models and many more, also aiopylgtv can be used to trigger it.

So, just keep every old firmware, LG can break anything anytime easily :)

ashlar42

8th September 2021, 21:18

chros

9th September 2021, 17:17

Instructions on Russian websites...
And in that post as well :)

does this apply to the 2021 models?
Probably, but with 2018 models this method is only required from v5.x.x, versions before we can use the SURE app method (also added info about it in that post) that also accepts USB and NSU versions.

aron7awol

10th September 2021, 15:04

I know you guys have a good amount of knowledge on 3dluts, so I'm asking this here...

I do not use a 3dlut currently on my OLED. What I would like to do is "emulate" Rec.709 color gamut with a 3dlut without tone-mapping to SDR. In other words, I'd like to watch HDR content, and toggle a 3dlut on/off with madVR, which simply reduces the color gamut to Rec.709, so that toggling it on/off shows the difference of WCG without changing anything else.

Do you know of a simple way to do this?

huhn

10th September 2021, 15:25

the current madVR version can't load a PQ to PQ with no tone mapping LUT so no not possible.

bt 2020 gamut to bt 709 is tone mapping so your request is kinda tough.

with tone mapping this would be trivial with nvidia.

quietvoid

10th September 2021, 15:53

You can just set madVR to map down to 709 while keeping the output HDR, that should work fine.
Not sure if madVR would still send it within a 2020 signal though.

huhn

10th September 2021, 15:58

the TV still thinks the signal is bt 2020 that way and it will now oversat.

except there is a new setting i don't know of.

aron7awol

10th September 2021, 16:02

aron7awol

10th September 2021, 16:06

I wouldn't be opposed to tone-mapping at the same time, as long as I'm able to toggle between two profiles/settings/3dlut where the only thing that changes is the gamut. Are you saying it would be possible if I tone-mapped to SDR BT2020 and then toggled a 3dlut to map to Rec709 within BT2020 primaries?

quietvoid

10th September 2021, 16:38

the TV still thinks the signal is bt 2020 that way and it will now oversat.

except there is a new setting i don't know of.
Usually there shouldn't be a problem with sending 709 within 2020 primaries container.
There are a couple threads on the color accuracy on AVS of the method but I don't really remember the details.

The only thing is to figure out if that's what madVR defaults to, or if the 2020 flag still needs to be enabled (NVIDIA).

I wouldn't be opposed to tone-mapping at the same time, as long as I'm able to toggle between two profiles/settings/3dlut where the only thing that changes is the gamut. Are you saying it would be possible if I tone-mapped to SDR BT2020 and then toggled a 3dlut to map to Rec709 within BT2020 primaries?I would assume the 3D LUT would either clip or do a slightly dumber mapping down to 709, so I'd avoid it.

You also don't need to tonemap to SDR, just tonemap in HDR space, but map down the colors only (the calibration config section).

aron7awol

10th September 2021, 19:40

Usually there shouldn't be a problem with sending 709 within 2020 primaries container.
There are a couple threads on the color accuracy on AVS of the method but I don't really remember the details.

The only thing is to figure out if that's what madVR defaults to, or if the 2020 flag still needs to be enabled (NVIDIA).

I would assume the 3D LUT would either clip or do a slightly dumber mapping down to 709, so I'd avoid it.

You also don't need to tonemap to SDR, just tonemap in HDR space, but map down the colors only (the calibration config section).
So, if I'm following what you're saying, you are saying to tell madVR the display is already calibrated to BT.709, but still output video in HDR format? Doesn't madVR ignore that calibration section when outputting in HDR format, and it only applies with SDR output?

quietvoid

10th September 2021, 21:45

So, if I'm following what you're saying, you are saying to tell madVR the display is already calibrated to BT.709, but still output video in HDR format? Doesn't madVR ignore that calibration section when outputting in HDR format, and it only applies with SDR output?

I haven't tried so I don't know yet :D
I see no reason why HDR output would be limited to WCG, at least I've used DCI-P3 instead of 2020 in the past.

chros

10th September 2021, 21:49

I initially thought it might be possible by creating a 3DLUT something like below. But I'm not sure if it's possible to do something like that while still outputting in PQ/HDR.

https://i.imgur.com/sPpmf0P.png
This should work! Just set gamma 2.2 instead of unmodified at "Tone curve", and you may even have to geneate a bt709 gamma 2.2 icm (not sure what the default gamma is). (madvr requires gamma 2.2 at 3dlut input, even with hdr10 output.)
You can create another one for DCI-P3 and switch between profiles. You can even measure primaries (around SDR nits range) to make them more "accurate". All this with hdr output.
When you do this, remove the factory 3dlut of the preset in the TV by uploading unity 3dlut. (Obviously you have to put this generated 3dlut into the bt2020 slot in madvr).
Btw, why is this interesting for you?

aron7awol

11th September 2021, 00:23

Btw, why is this interesting for you?
There is always debate in the PJ world about whether using a color filter is worth it to lose light in the process of increasing from something like 93% of P3 to 100% of P3. A conversation about that extended to someone saying that to them, the difference between 709 and P3 isn't all that noticeable. I disagreed with that, saying I feel like I can easily tell when there is WCG usage, but I've also never been able to toggle it on/off to actually see the differences on the fly on real scenes on the same exact display. I also feel like I could tell major differences in gamut coverage going from a Samsung KS8000 (with good coverage) to a Sony X900F (with poorer coverage), and then to my LG CX (with good coverage again).

Ideally, I'd be able to also toggle between full P3 and 90% P3 to test that particular comparison, but I figured comparing 709 vs P3 would be easier to set up. It still isn't as easy as I initially hoped. :)

quietvoid

11th September 2021, 01:20

I just tried it out and HDR output is indeed overriding the 709.

I guess the next easiest solution is to reencode your content down to 709.
At this point the LUT is probably simpler though.

aron7awol

12th September 2021, 01:33

I guess the next easiest solution is to reencode your content down to 709.
Is there an easy way you know of to reencode with a simple gamut clip to 709 while leaving everything else alone?

quietvoid

12th September 2021, 03:29

Is there an easy way you know of to reencode with a simple gamut clip to 709 while leaving everything else alone?

Not really, it's usually a couple software to setup. So without prior experience this isn't as simple.

I had the idea of trying with mpv and it seems to be a possible solution.
My setup was like this:
mpv.conf
target-peak=10000
d3d11-output-csp=pq
d3d11-output-format=rgb10_a2
tone-mapping=clip
profile=gpu-hq
gpu-api=d3d11
target-prim=bt.2020
target-trc=pq
format=default:colormatrix=bt.2020-ncl
video-output-levels=full
dither-depth=10
fullscreen=yes

input.conf
F3 no-osd cycle-values target-prim "bt.2020" "bt.709"

With this, I toggled between the primaries with F3 and saw noticeable desaturation for WCG red, at least.
However it seems to oversaturate most of the other colors..

chros

12th September 2021, 09:52

I created 2 PR for aiopylgtv (https://github.com/bendavid/aiopylgtv):
- Add support for dictionary via JSON in cmd line params (https://github.com/bendavid/aiopylgtv/pull/34)
- Add support for setting PC mode for an HDMI input (https://github.com/bendavid/aiopylgtv/pull/35)

Especially the second is really amazing, no need to click around and wait to switch between PC and non-PC mode! Although it takes a second or so, just like when switching into/out from Game preset.

I also created a basic version of it (without calibration related features) for command line usage (that runs only 1 function at a time), the reason was: it's faster about 1-2 seconds (due to it doesn't have to trigger that many request after connecting).
If someone is interested I can publish it properly, but here it is until then: bscpylgtv v0.1.0 (https://github.com/chros73/bscpylgtv):
- looks for the same ".aiopylgtv.sqlite" key config
- how to install it on Windows (https://forum.doom9.org/showthread.php?p=1946491#post1946491)

You can actually modify the greyed out settings as well, e.g. colorGamut in PC mode!

Examples:

bscpylgtvcommand 192.168.1.18 info_button
bscpylgtvcommand 192.168.1.18 set_current_picture_mode expert2
bscpylgtvcommand 192.168.1.18 set_current_picture_settings "{\"backlight\": 0, \"contrast\": 85}"
bscpylgtvcommand 192.168.1.18 set_current_picture_settings "{\"hdrDynamicToneMapping\": \"on\"}"
bscpylgtvcommand 192.168.1.18 set_current_picture_settings "{\"colorGamut\": \"auto\"}"
bscpylgtvcommand 192.168.1.18 set_pc_mode true hdmi2
bscpylgtvcommand 192.168.1.18 launch_app_with_params com.webos.app.softwareupdate "{\"mode\": \"user\", \"flagUpdate\": true}"
bscpylgtvcommand 192.168.1.18 power_off