madVR - high quality video renderer (GPU assisted)

[Asmodian]

Thanks! Even if it only improvements for calibration software support. ;)

madTPG expects the calibration software to always send TV levels = limited range. So basically HCFR has to send "black" as 16. This is just a communication rule between the calibration software and madTPG, though. madTPG uses the madVR settings to actually display those test patterns. So if madVR is configured to output PC levels, madTPG will output "black" as 0, just like madVR does. Which means that madTPG and madVR (+ any media player) should produce identical results.

Yes, I was trying to say that if you then view full range test patterns with MPC-HC that are not flagged full range you can see out of gamma values when using a 3DLUT (the white levels test pattern is a good example). :o

However, since this wasn't the case for matz I am at a loss. That test has always worked correctly for me.

[Sm3n]

wrong thread.
my bad :/

[SFX]

Is there any way to add calibration settings in the profile? My TV (PDP, Samsung 64F8500) displays 24p and 60p differently. And switch 3dlut manually is inconvenient.

[madshi]

Is there any way to add calibration settings in the profile? My TV (PDP, Samsung 64F8500) displays 24p and 60p differently. And switch 3dlut manually is inconvenient.
Allowing profiles for more settings is on my to do list, but it's more difficult to implement for devices, so I can't say how quickly I can implement that.

[SFX]

Allowing profiles for more settings is on my to do list, but it's more difficult to implement for devices, so I can't say how quickly I can implement that.

OK. I will wait for new releases. Thank you!

[huhn]

Really? madVR carefully present frames synced to the graph master clock. I don't see how madVR could produce any audio sync issue, regardless of any settings. FWIW, the user who reported this issue a few posts ago in the meanwhile found out that there isn't really any audio sync issue between Jinc and NNEDI3 for him, after all, and that the audio sync issue is limited to one specific video file.

some examples:

http://www.avsforum.com/forum/26-home-theater-computers/1357375-advanced-mpc-hc-setup-guide-110.html#post31116930

http://www.avsforum.com/forum/26-home-theater-computers/1357375-advanced-mpc-hc-setup-guide-109.html#post29866330

I personally have no issue with sync. every issue I have is in the bug tracker.

[makz]

Weird. How did you test this with MPC-HC+madVR? HCFR can only use madTPG to show test patterns, not MPC-HC+madVR. So you had to use different test patterns for MPC-HC+madVR, right?

So you get the correct/expected results when using madTPG? And incorrect results with MPC-HC+madVR? Is there any difference between enabling/disabling the 3dlut in MPC-HC+madVR? Maybe the 3dlut is simply not applied at all when using MPC-HC+madVR? Or is it applied, but somehow incorrectly? That is important to find out.



Weird. How did you test this with MPC-HC+madVR?
- AVSHD 709 patterns+MPC-HC+madvr+3dlut. Measurements were made with a laptop.

So you had to use different test patterns for MPC-HC+madVR, right?
- Yes, but 100% green in rec.709 space should not vary outside the generator and patterns AVSHD 709

So you get the correct/expected results when using madTPG? And incorrect results with MPC-HC+madVR? Is there any difference between enabling/disabling the 3dlut in MPC-HC+madVR? Maybe the 3dlut is simply not applied at all when using MPC-HC+madVR? Or is it applied, but somehow incorrectly? That is important to find out.
- with madTPG all very good. But MPC-HC + madvr - 100% oversatturated. 25,50,75 % correct. (Screenshots were posted)
Applying 3dlut noticeable, but the 100% color incorrect.
Edit Ranges does not change the chromatic. (In madvr i use 0-255. In mpc-hс + AVSHD709 2-Grayscale Steps patterns i see 16-235, as I understand it - all right )

[madshi]

some examples:

http://www.avsforum.com/forum/26-home-theater-computers/1357375-advanced-mpc-hc-setup-guide-110.html#post31116930

http://www.avsforum.com/forum/26-home-theater-computers/1357375-advanced-mpc-hc-setup-guide-109.html#post29866330

I personally have no issue with sync. every issue I have is in the bug tracker.
I can only guess that those issues are coming from something else. From a technical point of view, the way madVR works, it should not produce sync issues. Unless the user has chosen settings which result in a slide show. In that case madVR shows at least one frame out of 8 frames, to show anything at all. In this "slide show mode" sync can be lost. But as soon as the settings are changed to achieve fluid playback, sync should immediately be restored.

Yes, but 100% green in rec.709 space should not vary outside the generator and patterns AVSHD 709

with madTPG all very good. But MPC-HC + madvr - 100% oversatturated. 25,50,75 % correct. (Screenshots were posted)
Applying 3dlut noticeable, but the 100% color incorrect.
Edit Ranges does not change the chromatic. (In madvr i use 0-255. In mpc-hс + AVSHD709 2-Grayscale Steps patterns i see 16-235, as I understand it - all right )
Hmmmm... Does the debug OSD (Ctrl+J) say "limited range" or "full range"?

[makz]

Hmmmm... Does the debug OSD (Ctrl+J) say "limited range" or "full range"?
When I playing AVSHD 709 disk debug OSD say limited range

[6233638]

Really? madVR carefully present frames synced to the graph master clock. I don't see how madVR could produce any audio sync issue, regardless of any settings. FWIW, the user who reported this issue a few posts ago in the meanwhile found out that there isn't really any audio sync issue between Jinc and NNEDI3 for him, after all, and that the audio sync issue is limited to one specific video file.It may not be madVR's fault, but I've definitely had sync issues without the "delay playback" options enabled.
I have not tested it recently, but it seemed to mainly occur when going from windowed to fullscreen where the refresh rate changes.
I have madVR set to only change when entering full-screen, and use smooth motion when running in a window at 60Hz.

I tried disabling the option again back when I got a GTX970 because I really like how much more responsive it is when you're not waiting for 128 frames to queue up, but I ran into sync problems again.
I was never able to figure out if there was a specific delay or if it only happened some of the time with certain formats for example, but any time I don't have the "delay playback" options enabled, I end up running into sync issues that are corrected by re-enabling it.

[ryrynz]

Submitted a crash dump not long ago via MadVR, 0.87.14 is crashing in MPC-BE when subtitles are set to XYSubfilter upon playback of a file with subtitles, 0.87.13 doesn't crash.

[bcec]

Between NNEDI3@32 + Ordered Diffusion vs NNEDI3@16 + Error Diffusion 2, any recommendations on the combination I should pick for 1080p->2160p conversion? I guess the question is, which one has a more pronounced boost in image quality, the use of 32 vs 16 neurons in NNEDI3 or Ordered vs Error Diffusion.

[ryrynz]

I guess the question is, which one has a more pronounced boost in image quality, the use of 32 vs 16 neurons in NNEDI3 or Ordered vs Error Diffusion.

Definitely the extra neurons with NNEDI.

[bcec]

Definitely the extra neurons with NNEDI.

Thank you!

[Asmodian]

Definitely the extra neurons with NNEDI.

I agree, strongly. ;)

[octal9]

Submitted a crash dump not long ago via MadVR, 0.87.14 is crashing in MPC-BE when subtitles are set to XYSubfilter upon playback of a file with subtitles, 0.87.13 doesn't crash.having same exact problem with mpc-be (using several different builds) - had to revert to madvr 0.87.13 to prevent crash....

[anuswara]

Hi, I 'd linke to know if this madVR can fix my issue.
Monitor. 60Hz (unpossible to change this value!).
In both interlaced and progressive files I am unable to read the credits at the end, crawl.
no deinterlacing problems!

the source (but it happens with ALL files, regardless of the source)
http://www14.zippyshare.com/v/ZZYxU3Qi/file.html

what I really see:
http://www57.zippyshare.com/v/lFTwNYHY/file.html

to read the credits I have to pause the video, otherwise they are jumping.
I was wrong thinking that its due to 25 or 50fps on a 60Hz monitor! I have this issue in the crawls and in all the pannings with 30 or 60fps too.

I am skeptical to install additional filters perhaps my OS becomes unstable...so I ask if installing this madVR will fix my problem.
thanks a lot in advance. regards.

[Stereodude]

The joys of a sample and hold display. If you put your finger in front of the display and look at the end of your finger are the credits sharp as they roll by in the background?

[anuswara]

they are sharp, i.e. without blur
http://postimg.org/image/kzxmp15vj/

I am unable to explain :(
no blur, no jump, but they are pulsating i.e.
the I becomes II
the V becomes W
but as you see in the link above the deinterlacing yadif is very good. and this happens even with progressive sources!!!
so...with vlc, mpc-hc, powerdvd I am unable toread them, without pausing the playback.

no idea if its due to my monitor...surely not due to the files, because even the pannings with camcorder are smooth but pulsating.
hmmmm

[Stereodude]

they are sharp, i.e. without blur
http://postimg.org/image/kzxmp15vj/

I am unable to explain :(
no blur, no jump, but they are pulsating i.e.
the I becomes II
the V becomes W
but as you see in the link above the deinterlacing yadif is very good. and this happens even with progressive sources!!!
so...with vlc, mpc-hc, powerdvd I am unable toread them, without pausing the playback.

no idea if its due to my monitor...surely not due to the files, because even the pannings with camcorder are smooth but pulsating.
hmmmm
It's sample and hold induced "blur". It's a perception issue caused by the fact that the human visual system is basically incompatible with tracking motion on a sample and hold display (which LCD is). This is why LCD TVs have frame interpolation (120Hz or 240Hz processing) and strobing / scanning backlights. You will often find mention of "motion resolution" in reviews in TVs. 1080 lines being perfect performance for a 1080p display. It's a way to measure the perceived blur.

PS: This has nothing to do with madVR so we should probably take the discussion to another thread.

[anuswara]

Ok I warmly thank you for explanation! Much appreciated. Perhaps the same effect when you are in a car with a camcorder. The signal on the street (the limit of the speed 50km/h, or 70...). When the signal is far away you see in the video a perfect "50"; when the signal is near you, it starts to tremble (the more near you, the more tremble, obviously at the edge of the video; 50p offers less tremble than 25p ok! my video from DVB is 1080i25), so you are unable to read 50. Perhaps as you say this is a human eyes limit, because the issue arises in telefilms, where they use 65000dollars camcorders. Perhaps the solution is to have a monitor7tv with at least 120Hz (in Europe better 100 or 200Hz to avoid flickering). Ok sorry for my OT, I wrote here because people told me that installing madVR and selecting it as render in mpc-hc I could succesfully resolve my issues, but I seriously doubt hehe
Thanks again.
Regards

[yok833]

madVR v0.87.14 released

http://madshi.net/madVR.zip

* madTPG: fixed "madVR_Get/SetSelected3dlut" API
* madTPG: added "madVR_Enable3dlut" API
* madTPG: network search is a bit more extensive/agressive now
* fixed: madVR network functionality didn't always take WLAN into account
* target rect related change for DVBViewer
Again, only changes for calibration software (ArgyllCMS, Calman, LightSpace) and for DVBViewer. Maybe I'll find some more time for madVR development soon, but I can't say for sure right now...

Many thanks Madshi for the new release!
Do we need to recreate a new 3dlut after the last build?

[Asmodian]

No. The changes look to only make it easier for the calibration software to control the process. If you already have a good 3DLUT no changes are needed.

[yok833]

No. The changes look to only make it easier for the calibration software to control the process. If you already have a good 3DLUT no changes are needed.

Perfect thank you.. I will spend my times watching good movies instead :)

[Asmodian]

If anyone wants to provide feedback, corrections, and/or additions to my initial draft of a madVR rendering flow chart (http://forum.doom9.org/showthread.php?p=1710366#post1710366) please do. :)

Edit: After making the chart I again thought it would make more sense to use chroma upscaling instead of image upscaling to scale up the chroma when only doubling luma. It is, at least, easier to explain.

Ideally I could set a second chroma scaling option in image doubling, including NNEDI3, similar to the chroma scaling settings page. :D

As it is now I end up using Mitchell-Netravali for luma too (RGB really) anytime I am below the doubling threshold if I want to use it for both chroma scaling and scaling the chroma during luma doubling.

..but only if it wasn't very much trouble to implement. :)

[tfouto]

There will be a 64 bit version of madvr?

With HEVC decoding, 64 bit is really important...

[BetA13]

ok, just a heads up...
with the newest 347.52 Nvidia Driver, NNEDI3 gets artefacts when ENABLED (both in image douling and upscaling) with KEPLER GPU´s...

im not shure if only kepler GPU´s are affected..

hopefully nvidia fixes that..i really really really like NNEDI3....

greetz

[Anima123]

What exactly is the artifact looks like?

I have an 880M and didn't notice any visually dominant bug with NNEDI3. As I know, the dither option 1 do has bugs with recent drivers, which was already known and not fixed in the latest 347.52 driver.

[ryrynz]

with the newest 347.52 Nvidia Driver, NNEDI3 gets artefacts when ENABLED (both in image douling and upscaling) with KEPLER GPU´s...

No brainer really. Revert the driver.

[FireFreak111]

Well, news on the OpenCL front for Nvidia, 349.72 is coming, primarily for Windows 10 with the new WDDM 2.0 driver model support, DirectX 12 and, surpisingly, OpenCL 1.2.

http://www.neowin.net/forum/topic/1247418-wddm-20-nv-driver-released-via-wu/

OpenCL 1.2

Added OpenCl 1.2


From the wikipedia page of OpenCL 1.2:
DirectX functionality: DX9 media surface sharing allows for efficient sharing between OpenCL and DX9 or DXVA media surfaces. Equally, for DX11, seamless sharing between OpenCL and DX11 surfaces is enabled.

I can confirm via GPU Caps Viewer on Windows 10 with this driver that the driver supports OpenCL 1.2.

Does this mean the use OpenCL to process DXVA NV12 (Intel, AMD) option is now supported for Nvidia too? Or are we still waiting on something else on the Nvidia side?

Also, I can confirm the random 3-4 dot line, around 50 pixels long randomly flashing both vertically and horizontally with NNEDI3 enabled in the newest Nvidia drivers. Any form of NNEDI3 causes this.

[Asmodian]

ok, just a heads up...
with the newest 347.52 Nvidia Driver, NNEDI3 gets artefacts when ENABLED (both in image douling and upscaling) with KEPLER GPU´s...

im not shure if only kepler GPU´s are affected..


GK110 (or Titan at least) doesn't seem to be affected. At least I have not been able to notice anything watching few hours of 720p -> 1440 with NNEDI3 128 luma doubling. May be limited to the GK104.

Win 8.1 x64, v347.52, GTX Titan, X99, madVR v0.87.14

[cyberbeing]

ok, just a heads up...
with the newest 347.52 Nvidia Driver, NNEDI3 gets artefacts when ENABLED (both in image douling and upscaling) with KEPLER GPU´s...

im not shure if only kepler GPU´s are affected..
GK110 (or Titan at least) doesn't seem to be affected. At least I have not been able to notice anything watching few hours of 720p -> 1440 with NNEDI3 128 luma doubling. May be limited to the GK104.

I've not seen any NNEDI3 artifacts with my GTX 770 with any of the recent drivers on Win7, so it can't be all GK104 GPUs. Who else was able to reproduce this (GPU & OS)?

The images BetA13 posted earlier in the thread of his GTX670 look more like core or memory corruption from a bad GPU, rather than a driver bug. BetA13 are you positive your GPU is stable under maximum load? Can you run something like Furmark for 30 minutes without artifacts?

[tjcinnamon]

is it possible to save/export settings? I may be reformatting my HTPC at some point and I'm would like to save both my MPC-HC and MadVR settings.

If not, no biggie. I remember most of them. BUT it's nice to have a backup of my HTPC settings.

[BetA13]

I've not seen any NNEDI3 artifacts with my GTX 770 with any of the recent drivers on Win7, so it can't be all GK104 GPUs. Who else was able to reproduce this (GPU & OS)?

The images BetA13 posted earlier in the thread of his GTX670 look more like core or memory corruption from a bad GPU, rather than a driver bug. BetA13 are you positive your GPU is stable under maximum load? Can you run something like Furmark for 30 minutes without artifacts?

jup, i can confirm it is stable 30-40min Fur/OCCT, etc - benches, 2 runs of 3dmark, and 2 hours straight bf4.. i even tried downclocking since this was my first impression too. but i tested it now on 3 other PC´s with Nvidia, and her ein the thread where also people having the issue descriped..

like suggested, only reverting to the 344.75 help´s, wich also indicates driver Bug/error..
But, as i also do game with my PC, its hell for me..i dont want to change teh driver when i watch movies, lol..

i made a ticket to Nvidia, but i doubt it will change anything...maybe im wrong and they do listen, i dont know...

greetz

edit:
My system:

GTX 670, Win7 64bit...

[MokrySedeS]

is it possible to save/export settings? I may be reformatting my HTPC at some point and I'm would like to save both my MPC-HC and MadVR settings.

For madVR:
Simply copy your settings.bin file (located in madVR directory of course).

For MPC-HC:
Go to View --> Options... --> Miscellaneous
Click Export.
To restore simply double-click the saved .reg file.

[ryrynz]

For the sake of every non-German, you might wish to change Strg to Ctrl. ;D

Madshi could you edit the change log as well?

[Pat357]

I have a problem that I can't see or track my mouse movements in FSE mode.
There is no typical mouse "arrow" visible any more as soon as FSE is enabled.
When moving my mouse blindly, I see the typical MadVR FSE time-bar, and clicking starts seeking.
So this part seems to work, but I can't choose to what time I seek because the mouse is completely invisible.
It's obviously an annoying thing !

Environment Info :
MPC-HC (latest)
madVR v0.87.13
LAV v0.64
OS : Windows 7 (x64)
CPU: Intel Core i7 970 @ 3.20GHz (6C/12T)
GPU: GeForce GTX 680 (8 EU) @ 1084 MHz (347.25)

[jmac698]

I remember watching videos with jinc and antiring, the result was excellent! I no longer have a GPU capable of this, I was wondering if there's a way to render the same effect offline, basically as an upscaling filter, or is this pointless as there's better ways?

It was really such a pleasure to watch that way.

[Asmodian]

Jinc Resizer - Avisynth Plugin [v0.1.1] (http://forum.doom9.org/showthread.php?t=169813) but no anti-ringing. NNEDI3 is also a good offline option.

I have a problem that I can't see or track my mouse movements in FSE mode.
There is no typical mouse "arrow" visible any more as soon as FSE is enabled.
When moving my mouse blindly, I see the typical MadVR FSE time-bar, and clicking starts seeking.
So this part seems to work, but I can't choose to what time I seek because the mouse is completely invisible.
It's obviously an annoying thing !

Environment Info :
MPC-HC (latest)
madVR v0.87.13
LAV v0.64
OS : Windows 7 (x64)
CPU: Intel Core i7 970 @ 3.20GHz (6C/12T)
GPU: GeForce GTX 680 (8 EU) @ 1084 MHz (347.25)

I think this is probably a setting or player issue? This is not how madVR normally behaves.

All I can say is that madVR support IVideoWindow::HideCursor, so the media player should be in full control. Of course it is possible that there's a bug in madVR, but unless a media player developer tells me he's calling HideCursor and madVR doesn't react properly, there's nothing I can do. Except if all media players show the same problem for many users. That would be a sign that there's a bug in madVR somewhere...

[tjcinnamon]

I set the MadVR to change the display refresh rate to match the source (1080p60Hz -> 1080p24Hz) but it's not changing. I recently was watching at NTSC 29.97fps video but the display stayed at 60Hz.

The odd thing is, Kodi (XBMC) dvdplayer does change the refresh rate.

Any ideas,

GTX970, Older Samsung (C Series) LED, and can change display via other software players.

Thanks,
JOe K.

[Stereodude]

Did you force film mode?

[tjcinnamon]

Did you force film mode?

I found it under de-interlacing but it's still not working. Once I get home I'll be able to give you the settings. I am looking in the right place, correct?

[djsolidsnake86]

any chance to see madvr work with MPC 64bit?

[Asmodian]

madshi has said one day there may be a 64-bit madVR but not soon.

[tjcinnamon]

I set the MadVR to change the display refresh rate to match the source (1080p60Hz -> 1080p24Hz) but it's not changing. I recently was watching at NTSC 29.97fps video but the display stayed at 60Hz.

The odd thing is, Kodi (XBMC) dvdplayer does change the refresh rate.

Any ideas,

GTX970, Older Samsung (C Series) LED, and can change display via other software players.

Thanks,
JOe K.

Perhaps I'm looking for the wrong setting. Is this the setting that would help diminish Panning Smoothness issues?

[Stereodude]

I found it under de-interlacing but it's still not working. Once I get home I'll be able to give you the settings. I am looking in the right place, correct?
You have to force film and use the right decoding mode in LAV so that the IVTC routines in madVR kick in.

[tjcinnamon]

You have to force film and use the right decoding mode in LAV so that the IVTC routines in madVR kick in.

Sweet! it appears to be working!

[kazuya2k8]

ok, just a heads up...
with the newest 347.52 Nvidia Driver, NNEDI3 gets artefacts when ENABLED (both in image douling and upscaling) with KEPLER GPU´s...

im not shure if only kepler GPU´s are affected..


I seem to be not affected by this artefact. Enabled image doubling/chroma scaling NNEDI3. Tested on a 10-bit/8-bit 480p video into 1080p display. Using 347.52 driver with my GTX 660.

[bytheway]

Is there a current favorite GPU for using madVR at "maximum" that is also quiet (doesn't have to be silent) and has a reasonable TDP? Priority is excellent 23.976p/24p on a Sony TV that supports 1080p/24p. Thanks.

[ryrynz]

I think I'd put a vote in for the GTX 960 TBH, depends on your budget and power constraits though, pretty much everything out there is quiet enough.