madVR - high quality video renderer (GPU assisted)

[huhn]

Actually ringing is an artifact of sharpening of downscaled images.

For upscaling (chroma or luma) I don't think more than Bilinear is needed for accurate results.
Anyone disagree and why?

what?

[aufkrawall]

Works fine here (Windows 10 Insider). What OS and card do you have?
It works fine here for me too with Win 10 x64 final (latest updates) and R9 390.

[madshi]

the same way we choose with all other bigger upscaling factors
if we want to sharpen the source we use image enhacements
if we want apply the sharpen after upscaling use upscalling refinements
if we want use in both use in both
There's a reason I'm disabling the upscaling refinements for very small upscaling factors. As I said before: You don't like it, other users might like it. So if I fulfill your wish, other users might be unhappy. I can't make everybody happy without adding a new option. And that new option would be hard to understand.

as you suggest i tried again supersampling on v.19 and madvr still don´t activate upscalling refinements on my resolution movie = 720p, screen =1360x768
double checked on aways 2x supersampling and aways when upscaling is needed,
the image doubles, downscale but not apply the refinements
no image changes and no render times changes even if check all the refinements on high
Your posts are really hard to read. Try sorting your sentences into logical paragraphs, please, and don't use line breaks, except to separate paragraphs.

So is image doubling not activating for you, if you select "supersampling" in the image doubling section? Or does it activate, but upscaling refinement doesn't activate?

Just when I thought madVR can't get any better, and that it is already amazing as it is... this deringing thing is frigging fantastic! Ringing in the original bluray source is my number 1 annoyance. I am even annoyed by ones caused by very mild DNR+EE that is present in newer titles like Force Awakens. This thing is doing wonders! A big THANK YOU!
I haven't mentioned it yet but "reduce ringing artifacts" is great.. like all these new options it is hard for me to offer useful feedback for but it is a Good Thing. The updated version is better still.
Thanks a lot. :thanks:
:)

I have "delay playback start until render queue is full" turned on yet I still get the few dropped frames at the start of everything for a couple of secs.
Try with default settings, but "delay playback start" activated. Still the same problem?

Smooth motion FRC does need a moment to activate properly. That's just the way it is atm. It should only be a very short time, though, not a couple of seconds. Maybe your settings are too high so that the GPU has trouble keeping up? That might be partially responsible. Try lowering your madVR settings a bit to give the GPU more room to breathe. That might already fix the problem.

Is there a way for a 3rd party control system to control any of the MADVR functions via IP? I have currently been using Eventghost but I would like to eliminate the middleman and control it directly if possible.
Haven't we discussed this already a while ago? I'm pretty sure somebody asked a similar question not long ago, and I think it was you, wasn't it?

??? What's the point of this? Adding grain is just more noise isn't it?
Most movie sources have a bit of grain in them. The way the deringing works, it totally smoothes areas where ringing was detected. As a result these areas look smoother than the rest of the image, which is a visible artifact. So I'm adding grain to make these areas appear more natural and not overly smoothed over.

problem with the Sony HDR camp Demo
Just read a couple pages back. This was already reported and extensively discussed.

I understand your point but maybe it would makes sense to do a test build with different tresholds to play with? Maybe we will find a better compromise, maybe we won't.

I just tried it and I found more artefacts. Look at the teeth of the characters. It's maybe the only thing that keeps me to use it all the time with compressed videos.

Deringing 90.18 vs 90.19 (http://screenshotcomparison.com/comparison/173052)
Deringing off vs on (http://screenshotcomparison.com/comparison/173053)

Here's the original picture but I don't know if it will help because on my side, deringing doesn't seem to work on juste one picture.

http://postimg.org/image/5t877ur9t/
Oh well, I suppose I'll have to check where the artifact is coming from. Maybe I can fix it. I don't think modifying the thresholds it very promising. Maybe as a last resort, if everything else fails. But I'd much prefer not having a strength setting. The visible difference would be very small and users wouldn't know which setting to use.

As expected, unfortunately deringing in .19 looks quite dirty with cartoons.

.18:
http://abload.de/thumb/18enk21.png (http://abload.de/image.php?img=18enk21.png)

.19:
http://abload.de/thumb/1923j7l.png (http://abload.de/image.php?img=1923j7l.png)
It's not looking that dirty to my eyes, but my computer monitor isn't very good, I suppose. I guess I could try to adjust the grain level to the grain that is in the rest of the image. So I would add more grain for grainy sources, and less grain for very smooth sources.

@madshi

madvr 0.90.19 + mpc-hc (nightly) crash at start with last AMD crimson 16.5.2 hotfix

http://pastebin.com/XCWq7Akt
Does this one fix it?

http://madshi.net/madVR9019amdFix.rar

[James Freeman]

what?

Yeah.
A simple interpolation is all that it takes for upscaling.
You can't ADD detail that is not there, can you?

[Nevilne]

the irony of saying that in madvr thread

[madshi]

A simple interpolation is all that it takes for upscaling.
You can't ADD detail that is not there, can you?
OMG, are you serious? Why don't you take a nicely sharp photo, scale it down to 25% with a good downscaling algorithm, and then scale it back up with madVR to 100%. Then compare Bilinear to the more sophisticated algorithms. If you think that Bilinear is producing the best quality (or comes nearest to the original image), then you need to go visit an eye doctor, and very quickly.

[James Freeman]

No offense of course, just some speculation.
Interpolation is for interpolation which always starts with nearest neighbor, sharpening and anti-aliasing comes on top of that if wanted or needed.

[Ver Greeneyes]

Nearest neighbor, linear and bicubic (and its variants) are interpolation-based, but upscalers like SuperXBR and NNEDI3 work quite differently.

[aufkrawall]

It's not looking that dirty to my eyes, but my computer monitor isn't very good, I suppose. I guess I could try to adjust the grain level to the grain that is in the rest of the image. So I would add more grain for grainy sources, and less grain for very smooth sources.

Some areas look very dirty to me in comparison, e.g. the dark wooden wall at the stairway.
Well, looking forward to your grain detection. :)

OMG, are you serious? Why don't you take a nicely sharp photo, scale it down to 25% with a good downscaling algorithm, and then scale it back up with madVR to 100%. Then compare Bilinear to the more sophisticated algorithms. If you think that Bilinear is producing the best quality (or comes nearest to the original image), then you need to go visit an eye doctor, and very quickly.
mmd :D

[huhn]

Yeah.
A simple interpolation is all that it takes for upscaling.
You can't ADD detail that is not there, can you?

but you can remove them.

and bilinear is pretty good at making the image extremely unsharp so it is not that "accurate".

spline 3 is know to have pretty high psnr or was it gaussian?
what so ever PSNR is just math and not subjective.

i mean x264 can be used to achieve the highest possible PSNR but this usually doesn't result in a better subjective visible results. it is off by default for a reason.

doom9 was down so i guess i'm a little bit late.

[madshi]

No offense of course, just some speculation.
Why do you feel the need to speculate at all? It's so easy to test it. Please do yourself (and us all) a favor and do the test I suggested in my previous comment. If you do that, you'll quickly see that your speculation is way off.

[mark0077]

madshi, very small issue I have after every upgrade. I just want to get your opinion on it, as to whether to register it as a bug or not on madVR bug list.

After I upgrade versions, if I then launch a movie and go to fullscreen. If I then open the madVR properties via right click mpc-be window and open madVR properties, some sort of permissions dialog box appears behind the madVR window but I actually can't get to the "Edit Settings" section. Its like as if mpc-be and the "Properties" dialog are in some sort of modal setup where no combination of clicking Close / OK / Cancel / Alt-Tab can get me out of it. I have to kill with task manager or if I can't get to it which is the case right now, I have to reboot.

Just not sure if its anything that be made so that it doesn't appear on each upgrade.

EDIT: OK I reproduced in non full screen mode, and its Windows Smartscreen pop up thats saying its an unknown application. (Windows 10 64bit)

[James Freeman]

Nearest neighbor, linear and bicubic (and its variants) are interpolation-based, but upscalers like SuperXBR and NNEDI3 work quite differently.
This exactly the answer I was looking for, thanks.
I got confused about why would a good interpolation upscaler like Lanczos looks like a superscaling reconstruction filter (xBR) with very large upscaling ratios.
Still, I see ringing with anything besides bilinear with a special test pattern.

[ang3l]

Does this one fix it?

http://madshi.net/madVR9019amdFix.rar

yes thx :)

[nevcairiel]

Still, I see ringing with anything besides bilinear with a special test pattern.

Of course, but bilinear is much softer, ie. blurrier. Its all a matter of tradeoffs, especially when you stick to "normal" filters. Lanczos is sharper, but rings more.

Some people prefer sharper images, some people prefer the blurier, but cleaner look of softer interpolation.

There is no "correct" scaler.

[DragonQ]

Unfortunately no.

It's a new display every time? Does your AVR use a random serial number or how does that happen? Maybe it's just 2-5 different detected displays instead of an infinite number?
Not entirely sure but twice now I've had to reset it and there's at least 3 entries in the list of displays. :(

[madshi]

Not entirely sure but twice now I've had to reset it and there's at least 3 entries in the list of displays. :(
When that happens, simply drag & drop the new ID entry of the newly detected display to the original display. Afterwards you can remove the empty shell of the new display.

[ajp_anton]

Small request:

Current option:
"disable scaling if image size changes by only:"
Could you add a sub-option below it:
"allow integer-factor resizes to hit the same target"

to have for example a 1918x1078 video upscale to 3836x2156, for performance reasons.

[djfred93]

Yes, that is compression ringing (artefact from low video bitrate). However, the dering filter is more targeted towards downscale ringing, which are thin halos right beside lines, not ripples a fair ways away. The filter may not remove the ringing in this image very well/at all.

OK. Thanks. Unfortunately, you're right, the dering filter doesn't remove the ringing in this file.

[CarlosCaco]

Your posts are really hard to read. Try sorting your sentences into logical paragraphs, please, and don't use line breaks, except to separate paragraphs.

So is image doubling not activating for you, if you select "supersampling" in the image doubling section? Or does it activate, but upscaling refinement doesn't activate?





sorry about that, english is not my native language

if i set supersamplig image doubling activate but upscalling refinement doesn t activate

[Betroz]

@madshi, is there a way to compare super-xbr to NNEDI3 settings? Like super-xbr 100 gives the same image quality as NNEDI3 *** neurons?

It would be really nice to have some sort av list that compares the different settings, so people could have a general template to aim for according to the hardware capabilities their computers have.

[Razoola]

Where is image enhancing heading? Is anti-aliasing a thing of the past? What I'm saying is as resolution increases the requirements to achieve the best image change when upscaling. If your gaming in a 4k resolution on a 55inch screen is enabling any form of anti-aliasing simply putting uneeded work on the GPU for example? Of course upscaling is a little different but if your upscaling from say 720p to 1080p I feel using the same algo to upscale from 720p to 4k may not give the best result as using a different algo because of our eyes and the target resolution.

For me its all getting quite blurry in the gain some of these algos really bring. Probably mainly because I dont have the best eyesight in the world to be able to see the difference when sitting down to actually watch something at a normal viewing distance. On top of that of course not all content is the same.

But I do wonder if it is getting into the realm now where it should be possible for a renderer to look at the source resolution, destination resolution, and color spec. Then examine key frames in the movie and automatically pick the best scaling algo? For me personally I would enable such an option. Even if it required me waiting a minute for a movie to start while madVR calculated the best scaling based on the source material and output resolution etc, I think it would be worth it.

I guess a good start to that would be for a renderer to try a detect if the source material is a movie or animation by simply looking at frames of the source material only.

[ryrynz]

But I do wonder if it is getting into the realm now where it should be possible for a renderer to look at the source resolution, destination resolution, and color spec. Then examine key frames in the movie and automatically pick the best scaling algo?

You determine what you feel is "best" as everyone will have different opinions, BTW you can already set profiles to do to as you said.

I suspect it's likely not worth the effort to have preset profiles determined by graphics card power, while this has been mentioned a few times I don't think it will happen.

Plenty of guides online give you opinions on what to set based on general opinion, you just need to put some time and effort in to following them and determining what you think suits your display, taste & content.

[ionutm80]

Hi guys,

Really need your help here since I've done all possible things to solve the issue but to no avail.

My sytem chain:
- TV: Philips 32PFL9632, only accepts 60Hz resolution for PC input,
- HTPC: Win 10 x64, GTX 960 4GB, Intel Core I5 3470s, 8GB DDR3 1600, Intel SSD 128Gb
- Software: MPC-HC x64 nightly 207, LAV Filter 0.68.1, madVR 0.90.19, SVP 4 Pro

As stated above I'm using SVP to interpolate frames to 59.94 for my Blu-Ray rips 23.976 done with eac3to (thanks madshi for another great tool!). In madVR display modes I have input 1080p60, 1080p59 just to be sure. However madVR stats reports display at 59.93760 which is close to 59.94 but not a perfect match (I have another HTPC with an Intel HD4400 and I'm able to achieve 59.94006).I'm guessing this is the reason why I'm seeing "1 frame drop every 19.48 minutes", because otherwise rendering and present stats are more than ok and my rig should be more than capable to handle my madVR settings (chroma-up: Bicubic75AR, image-up: Lanczos 3 Taps + AR, downscalling: SSIM 1D, no doubling, ordered dithering, use D3D11, enable automatic fullscreen, everything unchecked in trade quality for performance). I've also tried to create a custom resolution with NVidia Control Panel of exact 59.94 but it seems that my TV is rejecting it, madVR starts showing 59.94 for 1 sec and than immediately switch back to 59.937 and sometime crashes.

Is there anything I can do about it? Is it my TV that only accept a fixed resolution, my NVidia which cannot produce a more accurate 59.94 or both?

I have also attached an image with madVR stats in full screen taken with my phone, sorry for the resolution.

http://i.imgur.com/IGVLpJS.jpg

The image is taken immediately after the movie started and I reset the initial dropped frames but I have also tested the movie for 1 hr and indeed after 20 minutes or so it started to dropped frames 15, 24, etc.

Thanks a lot in advance for any help.

[Razoola]

You determine what you feel is "best" as everyone will have different opinions, BTW you can already set profiles to do to as you said.

I suspect it's likely not worth the effort to have preset profiles determined by graphics card power, while this has been mentioned a few times I don't think it will happen.

Plenty of guides online give you opinions on what to set based on general opinion, you just need to put some time and effort in to following them and determining what you think suits your display, taste & content.

I am aware of this, just pointing out that one setting for say a 720p source movie may not work so well for another movie also at 720p given the filming techniques used by the director. You begin to get to the stage where the setting used for one does not work for the other.

[Asmodian]

I am aware of this, just pointing out that one setting for say a 720p source movie may not work so well for another movie also at 720p given the filming techniques used by the director. You begin to get to the stage where the setting used for one does not work for the other.

This is my experience as well, there aren't any universal settings optimal for all sources even at a single resolution, especially for upscaling refinement, artifact removal, and image enhancements.

However, having an algorithm pick optimal madVR settings is a far future concept. We don't even have a good way to measure video quality mathematically. ;)

The image is taken immediately after the movie started and I reset the initial dropped frames but I have also tested the movie for 1 hr and indeed after 20 minutes or so it started to dropped frames 15, 24, etc.

Your expected frame drop reported by madVR should only be a single frame every twenty minutes. It is to keep the video in-sync with the audio while usually playing the video correctly at the very slightly off frame rate. It sounds like your problem is somewhere else.

I suspect madVR interacting with SVP. Those are pretty light settings, is the GPU downclocking after a while?

Really need your help here since I've done all possible things to solve the issue but to no avail.

"All possible things" rules out pretty much everything we could suggest... What have you tried? :cool:

[ionutm80]

Your expected frame drop reported by madVR should only be a single frame every twenty minutes. It is to keep the video in-sync with the audio while usually playing the video correctly at the very slightly off frame rate. It sounds like your problem is somewhere else.

I suspect madVR interacting with SVP. Those are pretty light settings, is the GPU downclocking after a while?

"All possible things" rules out pretty much everything we could suggest... What have you tried? :cool:

Things tried:
- custom resolution NVidia / not working on my TV - starts with 59.94 but reverts immediately to 59.937
- lowered the settings in madVR to bare minimum / nothing changed
- on x86 chain I have 1 frame drop every 4.xy hours but only with ReClock / obviously not possible in x64
- checked on another system but with Intel HD4400 and there I can get very close to 59.94 --> 59.94006 / probably NVidia 60Hz resolution is in fact fixed at 59.937 ?

I do not know what else to try, but again my knowledge is limited and something escapes me ... or as you suggested maybe could be SVP linked meaning that probably I should interpolate to screen rate (much heavier on resources since more calculations needs to be done) instead of a fixed default 59.94 ...

Downclocking should be spotted with GPU-Z no? I can watch that to see if it's the case. Any suggestions on what else I should check?

[Razoola]

Hi guys,

Really need your help here since I've done all possible things to solve the issue but to no avail.

My sytem chain:
- TV: Philips 32PFL9632, only accepts 60Hz resolution for PC input,
- HTPC: Win 10 x64, GTX 960 4GB, Intel Core I5 3470s, 8GB DDR3 1600, Intel SSD 128Gb
- Software: MPC-HC x64 nightly 207, LAV Filter 0.68.1, madVR 0.90.19, SVP 4 Pro

As stated above I'm using SVP to interpolate frames to 59.94 for my Blu-Ray rips 23.976 done with eac3to (thanks madshi for another great tool!). In madVR display modes I have input 1080p60, 1080p59 just to be sure. However madVR stats reports display at 59.93760 which is close to 59.94 but not a perfect match (I have another HTPC with an Intel HD4400 and I'm able to achieve 59.94006).I'm guessing this is the reason why I'm seeing "1 frame drop every 19.48 minutes", because otherwise rendering and present stats are more than ok and my rig should be more than capable to handle my madVR settings (chroma-up: Bicubic75AR, image-up: Lanczos 3 Taps + AR, downscalling: SSIM 1D, no doubling, ordered dithering, use D3D11, enable automatic fullscreen, everything unchecked in trade quality for performance). I've also tried to create a custom resolution with NVidia Control Panel of exact 59.94 but it seems that my TV is rejecting it, madVR starts showing 59.94 for 1 sec and than immediately switch back to 59.937 and sometime crashes.

Is there anything I can do about it? Is it my TV that only accept a fixed resolution, my NVidia which cannot produce a more accurate 59.94 or both?

I have also attached an image with madVR stats in full screen taken with my phone, sorry for the resolution.

http://i.imgur.com/IGVLpJS.jpg

The image is taken immediately after the movie started and I reset the initial dropped frames but I have also tested the movie for 1 hr and indeed after 20 minutes or so it started to dropped frames 15, 24, etc.

Thanks a lot in advance for any help.

Why use SVP at all? Why not scrap SVP and simply enable madVR's smooth motion setting given it will make the required frames also? I think reclock will solve the situation for you given the slight timing situation you have.

[ionutm80]

Why use SVP at all? Why not scrap SVP and simply enable madVR's smooth motion setting given it will make the required frames also? I think reclock will solve the situation for you given the slight timing situation you have.

ReClock is only possible in x86 chain where indeed it allows me to have 1 frame drop every 4.xy hours, however I'm using everything in x64 ...

I know about smooth motion but that is frame blending and not frame interpolation. I kinda like SVP effect but this is a matter of taste.

My question towards more skilled people here is how close NVidia can be to 59.94? Because for me is 59.937. If I force everything at 1080p60 only in madVR I get 59.974 so I'm either below or above the target of 59.94. With Intel HD4400 I'm close to perfection 59.94006.

[Razoola]

Although reclock will make madvr report a better framedrop interval the value reported by madVR becomes invalid because reclock is constantly adjusting the audio so the framedrop never actually happens. Pity you are using x64, I keep 32bit just for reclock.

[Asmodian]

My question towards more skilled people here is how close NVidia can be to 59.94? Because for me is 59.937. If I force everything at 1080p60 only in madVR I get 59.974 so I'm either below or above the target of 59.94. With Intel HD4400 I'm close to perfection 59.94006.

Do not expect to get exact with Nvidia but madVR's report is not perfectly accurate either. madshi has said that you need to let it run and watch for frame drops if you want to tune the refresh rate perfectly.

[_rEuTeL_]

Hi madshi,

First of all thanks for all the good work :)
But could you please fix the bug with italic subs that breaks the 'move subs to bottom of the screen' option? It's the only issue I have with MadVR after months of use :)

Cheers

[huhn]

ReClock is only possible in x86 chain where indeed it allows me to have 1 frame drop every 4.xy hours, however I'm using everything in x64 ...

I know about smooth motion but that is frame blending and not frame interpolation. I kinda like SVP effect but this is a matter of taste.

My question towards more skilled people here is how close NVidia can be to 59.94? Because for me is 59.937. If I force everything at 1080p60 only in madVR I get 59.974 so I'm either below or above the target of 59.94. With Intel HD4400 I'm close to perfection 59.94006.

getting close to 59.94 is a waste of time you are still off to the audio clock.

if you want to stop repeating/dropping frames you have to match the audio clock.

the reason why intel has a "perfect" clock comes most likely from sharing the system clock with the video clock so they may have the same clock generator.

in the end it doesn't matter if the audio clock is different.

there are 3 clock in a system (maybe a lot more).
the system clock, the video clock and the audio clock.
the system clock is not really important in this case.

the clock deviation in madVR is important not the shown HZ.

[XMonarchY]

Damn, the new De-Ringer is quite heavy on my GPU. I had to lower SuperRes to 1 in both Image Refinement and in Chroma Upscaling for 21:9 content when using NNEDI3 Image Doubling. Still, I think it is worth it. I assume these specific ringing stuff in mostly in older films that were upscaled from SD to HD?

[ionutm80]

getting close to 59.94 is a waste of time you are still off to the audio clock.
if you want to stop repeating/dropping frames you have to match the audio clock.

Although reclock will make madvr report a better framedrop interval the value reported by madVR becomes invalid because reclock is constantly adjusting the audio so the framedrop never actually happens. Pity you are using x64, I keep 32bit just for reclock.

Ok, thanks for clarifying ... in fact what you both are saying is that w/o ReClock (only possible on x86 chain) I have no chance of having an accurate playback of 23.976 material on 60Hz screen even if I try to interpolate frames with SVP as close as possible to video clock because audio clock will not be synchronized, isn't it?
Are there any other methods except ReClock that would allow me to match the audio clock and still use everything in x64?

[huhn]

Ok, thanks for clarifying ... in fact what you both are saying is that w/o ReClock (only possible on x86 chain) I have no chance of having an accurate playback of 23.976 material on 60Hz screen even if I try to interpolate frames with SVP as close as possible to video clock because audio clock will not be synchronized, isn't it?
Are there any other methods except ReClock that would allow me to match the audio clock and still use everything in x64?

creating an custom resolution that is close to the audio clock.

madVR smoothmotion will take care of this too.

the sanear audio renderer has a planned feature like reclock.

[ionutm80]

the sanear audio renderer has a planned feature like reclock.

Isn't this the backend for MPC-HC Internal Audio Renderer?

[huhn]

Isn't this the backend for MPC-HC Internal Audio Renderer?

it is the new internal audio renderer in MPC-HC.
but this is a planned feature it is not implemented yet.

[Werewolfy]

Oh well, I suppose I'll have to check where the artifact is coming from. Maybe I can fix it. I don't think modifying the thresholds it very promising. Maybe as a last resort, if everything else fails. But I'd much prefer not having a strength setting. The visible difference would be very small and users wouldn't know which setting to use.


It would be great if you can get rid of this kind of artifact :) This issue happens each times a character smiles in this video.

[Stereodude]

if you want to stop repeating/dropping frames you have to match the audio clock.

the reason why intel has a "perfect" clock comes most likely from sharing the system clock with the video clock so they may have the same clock generator.

in the end it doesn't matter if the audio clock is different.
The R9 380 in my HTPC outputs the /1.001 framerates very accurately. I effectively don't get dropped or repeated frames from that. madVR generally shows somewhere >4 hours for the time per drop/repeat. It drifts around a little, but I don't have any issues. So, it's not just Intel.

[ionutm80]

The R9 380 in my HTPC outputs the /1.001 framerates very accurately. I effectively don't get dropped or repeated frames from that. madVR generally shows somewhere >4 hours for the time per drop/repeat. It drifts around a little, but I don't have any issues. So, it's not just Intel.

Then only the green team has a problem in achieving a very accurate clock ???

[huhn]

The R9 380 in my HTPC outputs the /1.001 framerates very accurately. I effectively don't get dropped or repeated frames from that. madVR generally shows somewhere >4 hours for the time per drop/repeat. It drifts around a little, but I don't have any issues. So, it's not just Intel.

this is luck because the audio clock can be totally different from system to system.

my r9 270 has an pretty good looking clock too but it doesn't match the audio clock what so ever.

[DragonQ]

When that happens, simply drag & drop the new ID entry of the newly detected display to the original display. Afterwards you can remove the empty shell of the new display.
Thanks, I've done that for all 3 detected display IDs. We'll see if it happens again.

[Warner306]

@madshi, is there a way to compare super-xbr to NNEDI3 settings? Like super-xbr 100 gives the same image quality as NNEDI3 *** neurons?

It would be really nice to have some sort av list that compares the different settings, so people could have a general template to aim for according to the hardware capabilities their computers have.

All of the algorithms will give you a different image. There is no value of super-xbr that looks exactly like NNEDI3.

The chart below was made some time ago to compare the different algorithms. I have no idea if it is still accurate today. It compares each algorithm using its default settings:

http://i1357.photobucket.com/albums/q760/Warner306/madVR%20Guide/Chart-Scaling-Algorithms_zps0wkmxshy.png

[baii]

There is not much reason to use x64 except you have problem with decoding performance, especially when reclock is needed. Madvr still does not benefit from x64 performance wise iirc?

Sent from my 306SH using Tapatalk

[QBhd]

When I made the switch to Windows 10, I also made the switch to an x64 workflow (since AMD now has that nice Custom Resolution section). There was a nice little boost in performance and I could bump a setting or two up. Now was that a Windows thing or an x64 thing? If I get a chance late tonight I will test this since I'm considering going back to using Re-Clock.

QB

[ryrynz]

Damn, the new De-Ringer is quite heavy on my GPU.

With just the first option ticked? It didn't make much of an impact on my 960 at all. What are you using?

[ryrynz]

Well, looking forward to your grain detection. :)


I think that would be a nice addition, but I have to wonder if perhaps some content might be slightly better served with a lower power dehaloing or perhaps a variation a little kinder to animation.

[zvans18]

With just the first option ticked? It didn't make much of an impact on my 960 at all. What are you using?

with 1080 movieed content on 1080 screen (nnedi3 16), it took a 12.xms average to ~15 with only first option on my 970

~33 to ~36 on 720 doubled and down to 1080, i could see needing to turn something lower if you're close to the edge

[Betroz]

All of the algorithms will give you a different image. There is no value of super-xbr that looks exactly like NNEDI3.

The chart below was made some time ago to compare the different algorithms. I have no idea if it is still accurate today. It compares each algorithm using its default settings:

http://i1357.photobucket.com/albums/q760/Warner306/madVR%20Guide/Chart-Scaling-Algorithms_zps0wkmxshy.png

Thanks for reply, but that was known to me :)