madVR - high quality video renderer (GPU assisted)

[madshi]

They seem full for me: (7-8), then frame dropped, then (0-8), and in 1ms (7-8) again, until next dropped frame.

Here is a log. Hope it will help. I double click the file which opens the player, go into fullscreen, reset stats (Ctrl+R), let the problems occurs (for 1 minute i guess), then pause, exit fullscreen, closes player.
http://dl.free.fr/jHUVQcG4y
Ok, I can see that every 5-10 seconds there's a "gap" in the log where nothing happens for about 100-150ms. Then suddenly everything works again, but those 100-150ms are enough to run the queues dry, which eventually results in a frame drop then.

Do you have any sort of application/service running in the background which might be using/blocking the GPU? Try closing any such application/service, and maybe also your browser. Does that help in any way?

Can you please create another log with this build:

http://madshi.net/madVRextendedLog.zip

YES it does!
Can we please try a 64 build?
Later.

[GCRaistlin]

madshi, can you please add the possibility to change the current device's refresh rate from the tray icon menu? Sometimes it's better to watch PAL videos at native 25 Hz, for example SopCast translations, sometimes not. It's easy to change playing speed in ReClock but not the refresh rate with madVR. So sometimes I have to uncheck "treat 25p movies as 24p" setting in madVR and then relauch the player - just to get the refresh rate I need now.

[detmek]

So here are 2 test builds.

@detmek:
http://madshi.net/madVR885Haswell.zip
Please try them all and let me know which fix the problem, if any. Thanks.



Hi. I tried all four versions and none of those fixes a problem. I used MPC-BE x86 Nightly for test.

I don't know if I mentioned before but I don't have dropped or delayed frames or glitches. Its just render queue is not full and frames are presented out of order.

The thing is, problem almost desapires if I turn of smooth motion or turn on FSE. It flickers just a 1-2 seconds. With FSE render queue goes full. And that is even with regular build. Without FSE and with Smooth Motion on it flickers all the time.

[aufkrawall]

So I'm somewhat torn now: With this PC game, chroma doubling helps, and also Jinc looks "different" to Catmull-Rom. So maybe I should put both chroma doubling and Jinc back in? But with a 4:2:0 source maybe there's no difference at all. However, "upscaling refinement" could result in sharpening the chroma channel so much that doubling after refinement might make a difference, once again. So maybe that's a good argument for putting it all back in? Opinions?

It's a very synthetic example, but it showed that with all the new sharpening possibilities, it'd be definitely nice if one could completely rely on a scaling algorithm that introduces is little aliasing as possible, and this applies only to Jinc (and NNEDI3).

Chroma doubling/quadrupling could be made accessible again. But since they are only useful in such rather synthetic cases, they don't need to be present where luma doubling/quadrupling options are, which are much more important. I think your initial explanation about users getting confused by it is still true.
However, I think "novice" users are confused by image doubling options in general anyway, so this isn't necessarily a real argument.

[pirlouy]

Do you have any sort of application/service running in the background which might be using/blocking the GPU? Try closing any such application/service, and maybe also your browser. Does that help in any way?
Han ! You're strong. :/
Indeed, it was GPU-Z.
I guess you don't need another log now. Sorry, I feel bad for not thinking about another application interfering. :/

I don't have any dropped frames/glitches now.

[tobindac]

So what's the current consensus about panels that are fed 10bit without dithering and give smooth output? Are they truly 10bit or something uknown in the pipeline dithers it? Or we just don't know?

[huhn]

So what's the current consensus about panels that are fed 10bit without dithering and give smooth output? Are they truly 10bit or something uknown in the pipeline dithers it? Or we just don't know?

the TV is most likely dithering in this case.

i pretty much made sure this is the case with my TV. either my TV is 10 bit and that not the case... or it is dithering the 10 bit input.

i know a way to dither the 10 bit input from madVR to 8 bit using the AMD driver but this is pretty much pointless a direct 16 bit to 8 bit dither is better.

[James Freeman]

To add to what huhun said;
At 8bit+ it is impossible to see the dithering pattern so it's a guess game.
One thing is for sure, a single dithering stage at the end is always better than several along the way, but again, at 8+ bit depth it's a guess game.
It even can be that this whole discussion is pointless if no one can actually see anything...
If one does not know or 100% sure that the panel is true 10bit, it is better to stick to 8+dithering.

Others claim that FRC in displays actually looks smoother because it is not dependent on the video frame rate like madVR is.
8 or 10, always keep dithering ON in madVR.

[madshi]

madshi, can you please add the possibility to change the current device's refresh rate from the tray icon menu? Sometimes it's better to watch PAL videos at native 25 Hz, for example SopCast translations, sometimes not. It's easy to change playing speed in ReClock but not the refresh rate with madVR. So sometimes I have to uncheck "treat 25p movies as 24p" setting in madVR and then relauch the player - just to get the refresh rate I need now.
I'm not sure how much sense that makes. I mean I could list the modes which are in the edit fields for display mode switching. But should I list more modes than that? I could enumerate all modes the GPU supports. But then you could also use the GPU control panel.

Hi. I tried all four versions and none of those fixes a problem. I used MPC-BE x86 Nightly for test.

I don't know if I mentioned before but I don't have dropped or delayed frames or glitches. Its just render queue is not full and frames are presented out of order.

The thing is, problem almost desapires if I turn of smooth motion or turn on FSE. It flickers just a 1-2 seconds. With FSE render queue goes full. And that is even with regular build. Without FSE and with Smooth Motion on it flickers all the time.
Well, then I'm lost. Without being able to reproduce this, I don't know what to do about it. FWIW, I might get a Skylake laptop sooner or later, maybe I can reproduce it then.

One last thing you could test (just for information): Does the same problem occur when using the "use a separate device for presentation" with D3D9?

Han ! You're strong. :/
Indeed, it was GPU-Z.
I guess you don't need another log now. Sorry, I feel bad for not thinking about another application interfering. :/

I don't have any dropped frames/glitches now.
No probs.

[Dogway]

The TV (a 2014 "H generation" Samsung so pretty widespread) turns automatically to 1080p60 mode to, as I said, reenable settings needed for 60p sources, like recorded TV programs, files played from USB, or HFR blurays or whatever, let's call this "Playback mode". In "PC mode" many settings are disabled and others behave distinctly (ie. sharpness setting is disabled at value 50 instead of 0). There's a third last mode if I got it right, "TV mode" for cable TV, etc, these 3 modes store TV settings independently (not exactly three but for simplicity let's assume this).

I just bought the TV 2 weeks ago, the first thing that happened when I plugged the HDMI was that my desktop was extending offscreen. There's a (known I guess) trick and it is to rename (or label IIRC) the HDMI input as PC. Only then I got my 1080 desktop within the TV limits, that is desktop matching the TV internal resolution without scaling, this is "PC mode". 1080p60 (as 1080p24) sends a different signal to the TV, the above mentioned "Playback mode", I can take pictures if you want but I doubt they could come useful at all. It does the expected thing if you write 1080p60 and your source is 60fps, but I didn't expect my display changing to 24p mode for 60fps sources when I only had 1080p24 in the box, that's why I asked for a profile group option for displays so I can enforce a display mode policy on only 24p sources. The reason to play 60fps content in PC mode was to circumvent an issue where BFI option is shared on the TV side between both modes 24p and 60p, not being needed on the later.

If I simply play video files from the HTPC with the madvr display mode box empty everything is fine, so I guess if you don't know what signal to send for PC mode the best option should be a passthrough.


edit: btw for some reason the statistics doesn't show chroma being upscaled with nnedi3, I have it set the same as luma (if > x2). It is instead using the upscaling settings in "upscale chroma", interestingly enough there's another NNEDI3 scaling option just below Jinc.

[madshi]

The TV (a 2014 "H generation" Samsung so pretty widespread) turns automatically to 1080p60 mode to, as I said, reenable settings needed for 60p sources, like recorded TV programs, files played from USB, or HFR blurays or whatever, let's call this "Playback mode". In "PC mode" many settings are disabled and others behave distinctly (ie. sharpness setting is disabled at value 50 instead of 0). There's a third last mode if I got it right, "TV mode" for cable TV, etc, these 3 modes store TV settings independently (not exactly three but for simplicity let's assume this).

I just bought the TV 2 weeks ago, the first thing that happened when I plugged the HDMI was that my desktop was extending offscreen. There's a (known I guess) trick and it is to rename (or label IIRC) the HDMI input as PC. Only then I got my 1080 desktop within the TV limits, that is desktop matching the TV internal resolution without scaling, this is "PC mode". 1080p60 (as 1080p24) sends a different signal to the TV, the above mentioned "Playback mode", I can take pictures if you want but I doubt they could come useful at all. It does the expected thing if you write 1080p60 and your source is 60fps, but I didn't expect my display changing to 24p mode for 60fps sources when I only had 1080p24 in the box, that's why I asked for a profile group option for displays so I can enforce a display mode policy on 24p sources. The reason to play 60fps content in PC mode was to circumvent an issue where BFI option is shared on the TV side between both modes 24p and 60p, not being needed on the later.

If I simply play video files from the HTPC with the madvr display mode box empty everything is fine, so I guess if you don't know what signal to send for PC mode the best option should be a passthrough.
Which mode is your PC in when everything works as intended in 60Hz with PC Mode? Seemingly it's *not* 1080p60, or is it? Otherwise it wouldn't harm if madVR switched to 1080p60. So your GPU must be in a different state than 1080p60, when PC Mode is working?

[GCRaistlin]

I could list the modes which are in the edit fields for display mode switching
Yes, this would be great.

[pirlouy]

So it did not change threw years... The only way to use "PC mode" with Samsung TV is through 60 (or 59) Hz, but no 50 or 24Hz.

[GCRaistlin]

you could also use the GPU control panel.
It's time consuming (to call Control Panel applet), and this way I won't get the original refresh rate back when the player will be closed.

[MS-DOS]

@everyone with the D3D11 minimize / Aero Peak problem:
http://madshi.net/madVR885minimize.zip
Does this one fix the problem?
Yes, thanks. I generally use a 64 bit one, though, would be nice to check it as well.
Render queues are still not full in windowed mode.

[James Freeman]

Yes, thanks. I generally use a 64 bit one, though, would be nice to check it as well.
madshi replied "Later" to that request, I guess 0.88.6 is coming shortly.
Render queues are still not full.
Same here, but I'm sure madshi will solve this one too.

[huhn]

So it did not change threw years... The only way to use "PC mode" with Samsung TV is through 60 (or 59) Hz, but no 50 or 24Hz.
no one in the industry cares about this issue and nearly all people that buy a TV doesn't even know what chroma subsampling is.
this is not a samsung only problem nearly all TVs have this problem.
the big exception are 1080p sonys they can do 4:4:4 at all refresh rates and most advanced setting still work.

but this is getting terrible off topic.

[tobindac]

the TV is most likely dithering in this case.

i pretty much made sure this is the case with my TV. either my TV is 10 bit and that not the case... or it is dithering the 10 bit input.

i know a way to dither the 10 bit input from madVR to 8 bit using the AMD driver but this is pretty much pointless a direct 16 bit to 8 bit dither is better.

I soon found out after that post you replied to that the AMD driver dithers by default. It can be turned off in the registry.

[huhn]

I soon found out after that post you replied to that the AMD driver dithers by default. It can be turned off in the registry.

it's good that AMD dithers 10 bit input to 8 bit if it is outputting 8 bit.

i have this registry "hack" active at the moment that's how i find out if my display is dithering and 10 bit was inputted in my screen.

but if the GPU is outputting 10 bit is is not dithering to 8 bit so this registry "hack" shouldn't be used for normal viewing.

[vivan]

Btw, can we get dithering for every frame presented? With Dx11 path and "present every vsync" option that should beat any FRC.

[James Freeman]

Btw, can we get dithering for every frame presented? With Dx11 path and "present every vsync" option that should beat any FRC.

Would you be kind and elaborate a bit?
I think currently madVR changes the pattern every video frame, so when I slow down the video the dithering pattern is also slowed down along with the video.
You want to say that the dithering pattern should change every refresh of the display itself and NOT the video right?

[detmek]

Well, then I'm lost. Without being able to reproduce this, I don't know what to do about it. FWIW, I might get a Skylake laptop sooner or later, maybe I can reproduce it then.

One last thing you could test (just for information): Does the same problem occur when using the "use a separate device for presentation" with D3D9?


Yes, it does! When I check "use a separate device for presentation" with D3D9 I get exactly the same problem. Can it be fixed (if the cause is the same)?

[x7007]

I guess it's a known issue with Madvr 88.4 and 88.5 because I have an issue watching any movie with some random presention error and Renderer queue and all others are not full.

What is the differences between Font Renderer : Vector Text Renderer and Bitmap Text Renderer ?

For example the vector with madvr 88.4-88.5 works ok with the known renderer issues, but as soon I change to Bitmap I have drop in frame every couple time a new subtitles sync is shown.

[ibius]

Hi, my first post here in madVR thread.
madshi, thank you very much for this great piece of software, I love it, I love SM FRC on my 60Hz screen and I love all the other bells and whistles it gives me.
I'm sooo used to madVR now, can't use EVR(CP) anymore.

I like the (small) performance increase D3D11 presentation path provides on my system, and it also being the requirement for 10-bit output, looks like we/I will have to switch to D3D11 sooner or later, so I wanted to give you my feedback on it as well.
D3D9 works perfectly for me, D3D11 needs some work.

Issues with D3D11:
1. Going FSE is fine, but after first seek presentation glitches start piling up by the dozen every second during playback, enabling 'Present a frame for every v-sync' fixes it, but windowed FS mode works fine without. All queues apart from render queue (see below) are full at the time.
2. After going windowed FS or FSE, render queue is always only 3-5/8, other queues are full, and I get increased CPU usage, e.g. from ~7% to ~30% on my 3,6GHz Intel Core2Duo.
3. Sometimes after seeking/stopping playback video gets stuck in a loop, jumping between few frames.
4. Render times are better compared to D3D9, but present times are way worse, e.g. going from 0.12ms to 0.9x/1.xx ms in windowed mode, in windowed FS or FSE they are even higher (although only chroma is scaled).


Config:
latest nightly madVR, LAV, MPC-HC x64, GTX 550Ti (2 different stable Fermi-drivers tested) on 60Hz 8-bit Eizo, Win8.1 x64

Flush settings all set to 'Dont flush', except for 'after last render step' only when checking render times, never had tearing, dropped frames or presentation glitches (with D3D9).

'use a separate device for presentation' enabled, it gives me better render times.

windowed(FS) mode (new path), FSE only used to test D3D11.

SM FRC enabled, delay playback start/after seeking too enabled, ordered dithering, few of the default trade quality for performace options enabled, no new scaling refinements enabled (yet).

LAV Video set to defaults, using DXVA2-CBd decoder

Nvidia control panel set to defaults, including 'Application-controlled' for 'Max. pre-rendered frames'

[tobindac]

I noticed that ordered dithering is a very noisy process. I found it fun lately to turn on 1bit output and see very obviously some algorithmic patterns. It appears not dithering at all gives the most "stable" unnoisy picture while error diffusion is more stable and 'random' is a total mess.

I'm surprised, but it appears no dithering at all is a viable choice.

Choices..

[huhn]

I noticed that ordered dithering is a very noisy process. I found it fun lately to turn on 1bit output and see very obviously some algorithmic patterns. It appears not dithering at all gives the most "stable" unnoisy picture while error diffusion is more stable and 'random' is a total mess.

I'm surprised, but it appears no dithering at all is a viable choice.

Choices..

if you like banding yes it is a viable choice.

[tobindac]

if you like banding yes it is a viable choice.

Spare me the passive aggressiveness. If the source is suitable, it may be better in some specific cases. I know banding makes it worse in some cases.

I wouldn't be surprised if various modern movies from good sources do not require dithering filters for a good viewing experience.

If that is achieved why not avoid the extra noise? The packaged result might be a better outcome.

[Ver Greeneyes]

Yes, dithering is the process of introducing noise to compensate for rounding errors. This helps average out the rounding errors either across the image or between frames (or both), and the noise is based on the actual error being compensated for, but it's still noise.

[tobindac]

Yes, dithering is the process of introducing noise to compensate for rounding errors. This helps average out the rounding errors either across the image or between frames (or both), and the noise is based on the actual error being compensated for, but it's still noise.

It reminds of antialiasing. It smooths the picture, yes. But at the end of the day it's even with the best algorithm a glorified version of "targeted blur". :)

[Ver Greeneyes]

All the processing that madVR does (including the color space conversions and the chroma doubling that almost all material needs) introduces rounding errors, so just because the source is 'good' doesn't mean you don't need dithering. It might mean you don't need debanding, but that's a different matter.

But - sure, if the noise bothers you when madVR is set to the correct output bitdepth, turn it off. Or turn off "change dither for every frame", which will make it less effective but might make it less visible (except on very uniform surfaces that are displayed for a considerable length of time). But it does mean you'll get more of a rounding error on average, reducing color accuracy.

[huhn]

Spare me the passive aggressiveness. If the source is suitable, it may be better in some specific cases. I know banding makes it worse in some cases.

I wouldn't be surprised if various modern movies from good sources do not require dithering filters for a good viewing experience.

i can't disagree that not all scene need dithering no question.

but a movie without a dark scene or gradations is unlikely.
even with really high bit deep there could be banding in a BD when dithering isn't used.

[nevcairiel]

I wouldn't be surprised if various modern movies from good sources do not require dithering filters for a good viewing experience.

You will *always* need dithering, because its about the mathematical process, not about the quality of the source. In fact, its entirely independent of that (although its going to be more obvious in some sources than in others, of course)
If such a perfect source would exist, then dithering also wouldn't add noise, since dithering, by definition and design, is only used to spread the rounding error to multiple pixels, instead of accumulating it all in one. If there wouldn't be a rounding error, there would be no "noise".

At the end of the day, movies are all going to be in YUV, and YUV needs to be converted to RGB. This conversion process (if done right) results in floating point values, which need to be converted to 8-bit (or recently, 10-bit) on output. If you don't use dithering, you will effectively cut off data. Thats also were your analogy goes wrong, dithering isn't about hiding flaws, its about preserving data when converting to a lower bitdepth (ie. from 16/32-bit float, to 8/10-bit integer for output)

Of course you are free to use whatever settings you like, but please don't spread them as advice.

[Dogway]

I'm a bit confused when you ask "what mode my PC is", where should I look that? Control panel just said (roughly) 1080 native, 60Hz.

Anyways I had a look and I don't find any video that turns my TV to 1080p60 mode, it stays always in PC mode which is fine and bad at the same time... I would expect a bit more consistency for these things (I should be able to treat 60fps sources as every other non-PC source, TV settings wise). 1080p59 is also defaulting to PC mode.

Then I played a 50fps video and screen turned to 50p mode, so here the story does apply.
Anyways I removed the 1080p50 entry and it defaulted to 1080p60 now seemingly and oddly enough "PC mode" (I still think passthrough should be a better idea?). I will report back when I find something strange. Sorry to disturb.

I'm still curious on why I can't seem to enable chroma NNEDI doubling. I even tried with "always - if upscaling is needed", only luma gets upscaled with nnedi.

[cvrkuth]

Yes, madvr display modes work correctly now. Thanks!
My playback is fine (no dropped frames, no glitches).

Anyway, just for info .. when I watch movie 24.000 and refresh rate fix is active, render queue and present queue are low (render queue 3-5/8 and present queue 2-3/8. Average stats present 31.00 ms which is abnormally high, usually 0.05 ms.)
When I watch movie 23.976 and refresh rate fix is not active, everything is fine with numbers (render queue 7-8/8 and present queue 6-7/8. Average stats present is 0.05 ms.)

Hope this help,
Regards

[huhn]

I'm a bit confused when you ask "what mode my PC is", where should I look that? Control panel just said (roughly) 1080 native, 60Hz.

Anyways I had a look and I don't find any video that turns my TV to 1080p60 mode, it stays always in PC mode which is fine and bad at the same time... I would expect a bit more consistency for these things (I should be able to treat 60fps sources as every other non-PC source, TV settings wise). 1080p59 is also defaulting to PC mode.

Then I played a 50fps video and screen turned to 50p mode, so here the story does apply.
Anyways I removed the 1080p50 entry and it defaulted to 1080p60 now seemingly and oddly enough "PC mode" (I still think passthrough should be a better idea?). I will report back when I find something strange. Sorry to disturb.

I'm still curious on why I can't seem to enable chroma NNEDI doubling. I even tried with "always - if upscaling is needed", only luma gets upscaled with nnedi.

nnedi3 chroma doubling was removed under image doubling.

[Dogway]

nnedi3 chroma doubling was removed under image doubling.

Since what version? I'm using v0.87.21

[huhn]

Since what version? I'm using v0.87.21

since 88.x

but if you are using 87.21 can you give me a screen from the osd ?

[Dogway]

since 88.x

but if you are using 87.21 can you give me a screen from the osd ?

http://i.imgur.com/HcNLs3Ot.png (http://i.imgur.com/HcNLs3O.png)

http://i.imgur.com/IP5n9PHt.png (http://i.imgur.com/IP5n9PH.png)

[huhn]

on this screen chroma doubling is used.

BTW. any reason you didn't use smooth motion on this 60 Hz screen?

just interested in your reasoning.

[Asmodian]

I noticed that ordered dithering is a very noisy process. I found it fun lately to turn on 1bit output and see very obviously some algorithmic patterns. It appears not dithering at all gives the most "stable" unnoisy picture while error diffusion is more stable and 'random' is a total mess.

I'm surprised, but it appears no dithering at all is a viable choice.

Choices..

No dither at all is NOT a viable choice.

I am not at all sure how one could conclude dithering is bad when watching 1-bit output. Watching 1-bit output without dithering makes it hard to see what is going on. Almost all detail is lost. Turn off "change dither for every frame" if you want a more stable image, DO NOT turn of dithering entirely. Ordered dithering has very low apparent noise even at only 4-bit (256 colors), especially with "change dither for every frame" off.

But - sure, if the noise bothers you when madVR is set to the correct output bitdepth, turn it off. Or turn off "change dither for every frame", which will make it less effective but might make it less visible (except on very uniform surfaces that are displayed for a considerable length of time). But it does mean you'll get more of a rounding error on average, reducing color accuracy.

"change dither for every frame" doesn't make dithering less effective or less accurate. It makes dither patterns easier to see but average rounding error and color accuracy is the same with or without "change dither for every frame".

Turning off "change dither for every frame" trades more visible dither patters for less apparent dither noise.

[x7007]

My playback is fine (no dropped frames, no glitches).

Anyway, just for info .. when I watch movie 24.000 and refresh rate fix is active, render queue and present queue are low (render queue 3-5/8 and present queue 2-3/8. Average stats present 31.00 ms which is abnormally high, usually 0.05 ms.)
When I watch movie 23.976 and refresh rate fix is not active, everything is fine with numbers (render queue 7-8/8 and present queue 6-7/8. Average stats present is 0.05 ms.)

Hope this help,
Regards


Same happens to me

[Dogway]

on this screen chroma doubling is used.
BTW. any reason you didn't use smooth motion on this 60 Hz screen?

How do you know that? shouldn't it be "chroma > nnedi16 > Mitchel..."?
I never use smooth option (a gimmick IMO) but this is my PC monitor not my TV anyways.

[Ver Greeneyes]

"change dither for every frame" doesn't make dithering less effective or less accurate. It makes dither patterns easier to see but average rounding error and color accuracy is the same with or without "change dither for every frame".

Turning off "change dither for every frame" trades more visible dither patters for less apparent dither noise.Changing the dithering pattern every frame is more effective because it adds a time-based element. Let's take Error Diffusion as an example with a hypothetical losslessly compressed test pattern showing large uniform patches of color. Without temporal dithering, each pixel is going to have a particular color all the time - rounded either up or down depending on its position. Only the patch taken as a whole will have the right color on average. With temporal dithering, each pixel will have the correctly rounded value on average.

Of course in real material the difference is going to be negligible because of changes in what's being displayed, and noise in the source material subtly changing the dithering decisions each frame. Still, if you combine temporal dithering with something like smooth motion so that it changes for every screen refresh, it makes it much more effective (which is why I would prefer dithering to be tied to vsync rather than the video frame rate - but I always use smooth motion anyway, so it's a moot point).

[huhn]

How do you know that? shouldn't it be "chroma > nnedi16 > Mitchel..."?


the first chroma line is always the 4:2:0 to 4.4:4 part and nothing else.
and because there is no second chroma line image doubling is doing this.
I never use smooth option (a gimmick IMO) but this is my PC monitor not my TV anyways.

why gimmick?

please don't compare it to frame interpolation like SVP or the interpolation from TVs.
but usually it is not needed on a TV anyway.

[Dogway]

the first chroma line is always the 4:2:0 to 4.4:4 part and nothing else.
and because there is no second chroma line image doubling is doing this.

I think I understand what you mean, first line is converting to 4:4:4, then it upscales the whole thing with nnedi (hence image >, and not luma >), but in my opinion it shouldn't be doing that... it's not a good idea to use a a scaler before nnedi because you are amplifying artifacts, also you are adding one more step to chroma.

chroma > mitchell > nnedi (double) > resizer (for up or downscale to 1080)

has the behavior changed in last version?

edit: btw I don't think what you say is right, I disabled chroma nnedi and I get the same OSD output.

[Asmodian]

Changing the dithering pattern every frame is more effective because it adds a time-based element. Let's take Error Diffusion as an example with a hypothetical losslessly compressed test pattern showing large uniform patches of color. Without temporal dithering, each pixel is going to have a particular color all the time - rounded either up or down depending on its position. Only the patch taken as a whole will have the right color on average. With temporal dithering, each pixel will have the correctly rounded value on average.

I see the point, however this temporal effect isn't going to give "correct" pixels values because what value the pixel gets over time is random, not influenced by accumulated error. madVR's dithering is spacial dithering, not temporal, and the option does not change this.

[Anime Viewer]

Issues with D3D11:

3. Sometimes after seeking/stopping playback video gets stuck in a loop, jumping between few frames.


I noticed that too (when pausing videos), but wasn't sure how to word it. The best description I could think of at the time was it jumping back and forth between frames during pause. I've found that if you change scalers (up or down) it stops it and stabilizes the picture, so it may be a combination of D3D11 and scaler coding/changes.

[vivan]

I see the point, however this temporal effect isn't going to give "correct" pixels values because what value the pixel gets over time is random, not influenced by accumulated error. madVR's dithering is spacial dithering, not temporal, and the option does not change this.It will, because the chance the pixel will be rounded up (or down) is proportional to it's remainder value.
The way it changes is random - so it's random temporal dithering. It might be possible to make it ordered (imagine 3D matrix for ordered dithering, which is "good" in all dimensions).

Ordered dithering works on probability too, the only difference from random is that it uses special "good" matrix. "Good" means that peaks and valleys are as far from each other as possible.

[Asmodian]

I think I understand what you mean, first line is converting to 4:4:4, then it upscales the whole thing with nnedi (hence image >, and not luma >), but in my opinion it shouldn't be doing that... it's not a good idea to use a a scaler before nnedi because you are amplifying artifacts, also you are adding one more step to chroma.

chroma > mitchell > nnedi (double) > resizer (for up or downscale to 1080)

has the behavior changed in last version?

edit: btw I don't think what you say is right, I disabled chroma nnedi and I get the same OSD output.

Remember the chroma is half the resolution of the luma so you need to upscale it before doubling somehow, an extra step for chroma is unavoidable. If you don't think it is a good idea to use mitchell before NNEDI3 don't tell madVR to do that. You can use NNEDI3 to do this first upscale as well (set it in 'chroma upscaling').

On my system with v0.87.21 if I disable chroma doubling I get a second chroma line below luma that says chroma is being upscaled with what is set in 'image upscaling' (and 'image' changes to 'luma'). I don't get the same OSD output if I disable chroma doubling.

[Werewolfy]

Stupid me. I forgot that the refresh rate fix also needs new madHcNetXx.dlls. You'll find them here:

http://madshi.net/madVR885refreshRateFix2.zip

These files together with the modified madVR(64).ax should make the refresh rate work again.


Your fix never worked for me even before but this time it works! Thanks, great work.