madVR - high quality video renderer (GPU assisted)

[Anime Viewer]

Using it on your 8-bit monitor will give you worse quality since it will basically bypass madVR's dithering to fall back on the dithering of your display or GPU.

Are you sure?
I think switching to 10bit in madVR does not disable dithering, but it does make the dithering less strong.
If madVR would have 16bit option with dithering enabled, on an 8bit monitor it would look completely undithered.

Madshi wrote about it in another post. The way he currently has things set is if 10-bit is active it will switch dithering mode from Error Diffusion (if one of the Error Diffusion options is selected for dithering) to Ordered Dithering. He wrote that at that depth no significant visual improvements can be seen using Error Diffusion on 10-bit screens and the resources used for Error Diffusion were wasted, and could be better utilized elsewhere (I agree).

The old test here (for those not familiar with it) is to go into madVR display settings and change the bit-depth for the monitor to anything from 1 to 4 bit (which will make the dithering effects visible), and then toggle through the different dithering patterns to find which effect(s) you like the look of best. The difference between Error Diffusion and Ordered Dithering is pretty insignificant. When you change your madVR display setting back to 8 or 10-bit depth the patterns should completely disappear (making the choice even more insignificant).

There's a new superresolution upscaler for still images: https://github.com/nagadomi/waifu2x/ based on http://arxiv.org/abs/1501.00092

There also is a webservice running that implementation which has been trained with anime-style images and provides very good results on that type of material, especially on fine lines that are barely a pixel wide and fairly aliased. It also seems to introduce fewer fake details on flat surfaces than nnedi.
If you run it on real life images you'll get oil paintings.
Specialization a specific kind of content seems to have a pretty big impact.

Source image (http://i.imgur.com/uWMu9zt.png)
new upscaler w/ noise reduction (http://i.imgur.com/WnHLPE4.png)
new upscaler w/o noise reduction (http://i.imgur.com/QG8aVtr.png)
NNEDI3 32 w/ SuperRes (http://a.pomf.se/ywwfia.png)
NNEDI3 64 w/o SuperRes (http://a.pomf.se/smpgig.png)


So NNEDI3 blows waifu2 away in the most important area (line construction) check out left side hair clip (nearest neighbour upscale? eww) and the hair outlines.
Sure waifu2 used with a blur is cleaner and sharper but it's totally overdone.

I agree with 8472. I like the look(s) of waifu2 a lot better than NNEDi3.

I use DisplayPort. I guess its just my monitor that does not accept 10bit. I wonder if there is a way to force it... I mean there is DEFINITELY dithering going on on my monitor. I can see it easily.
Someone already mentioned it, but it would not do any good forcing your system to output 10-bit if your monitor/screen/tv is limited to 8-bit. Like previously stated it can make your image worse.
Perhaps your source video is the problem. If the source/video is lower bit than the screen you're displaying on then you're going to get a worse quality picture/video than you would if the source matched your screen bit depth. Having a 10-bit source is ideal for displaying 10-bit mode on a 10-bit screen. If you think you can see dithering with your madVR settings set to 8 or 10-bit for your screen do the bit changing test I posted above. Toggle through each setting from 1-bit all the way to 10-bit in your madVR settings, and see how the dithering effects become less and less noticeable. Once you get to 5 or 6-bit you shouldn't be able to see the dithering pattern/effects any longer. Still think you're seeing dithering? Its far more likely that you've got a bad/noisy/grainy/garbage filled source or your scaling choices are adding artifacts to the video.

[ryrynz]

I agree with 8472. I like the look(s) of waifu2 a lot better than NNEDi3.

There's only one good thing about waifu2 IMO and that it's line detection code. Bumping up NNEDI3 to 256 neurons and adding a sharpener and denoiser of your choice would produce a superior image overall.

You really can't compare 64 neurons without any sharpening to an image that's undergone sharpening with or without denoising. Of course some people are going to prefer an over sharpened image.. I'm not one of them.

I've started a thread about it here (http://forum.doom9.org/showthread.php?p=1722931#post1722931).

[Shiandow]

I wonder if shiandow could adapt NNEDI3 with waifu2x's improved neural guesswork.

As far as I can tell they're not using better "guesswork" they're just using far far more neurons, they also look at a larger region of the image to calculate a single pixel (hence why it's better at removing aliasing). I estimate it's at least 20x slower than NNEDI3 with 256 neurons, but even that is a very generous estimate, it could easily be 100x.

[ryrynz]

I estimate it's at least 20x slower than NNEDI3 with 256 neurons, but even that is a very generous estimate, it could easily be 100x.

Pretty much ends the discussion of comparisons right there, does make me wonder what NNEDI3 with 512 neurons would look like though.

[x7007]

I succeed to use the 10bit , but my render and present queue are 2 and 1.

Without present a frame for every Vsync I had tons of presentation errors / with it none

Any idea how to fix the render and present half queues ?

Does watching movies in 10bit will be any usefulness ? what will it be good for if not for movies

[James Freeman]

I was right about 8bit + FRC NOT being the same as true 10bit. True 10bit monitor would be allowed a setting of 10bit in nVidia CP. Using 10bit setting in madVR is a bad idea for 8bit + FRC monitors.

You are absolutely WRONG about everything.

I have a Dell U2410 which is 8bit+FRC and it can receive 10bit signal easily, in fact it is what I use now in Windows 7 by default.
I can select between 8 or 10bit in nvidia control panel, I use DisplayPort.

Using 10bit setting in madVR is a GREAT idea for 8bit + FRC monitors.
For the simple reason that FRC dithers in the refresh rate of the display, but madVR is tied to the frame rate of the video played (until of course madshi decides to change that, if ever).
And because madVR does not dither the perfect integers of 10bit output but 8bit+FRC display is, so not double dithering, everything falls into place like magic.

Yes, if one has a 8bit+FRC or true 10bit display that accepts 10bit signal, please use madVR in 10bit+Dithering with FSE and D3D11, you should get a better/cleaner picture (although practically invisible).
Also keep windows in 10bit so that the switch to FSE is faster.

[huhn]

You are absolutely WRONG about everything.

I have a Dell U2410 which is 8bit+FRC and it can receive 10bit signal easily, in fact it is what I use now in Windows 7 by default.
I can select between 8 or 10bit in nvidia control panel, I use DisplayPort.

Using 10bit setting in madVR is a GREAT idea for 8bit + FRC monitors.
For the simple reason that FRC dithers in the refresh rate of the display, but madVR is tied to the frame rate of the video played (until of course madshi decides to change that, if ever).

but if he can't output 10 bit to this screen it is doing nothing positive.

[James Freeman]

but if he can't output 10 bit to this screen it is doing nothing positive.

Correct.
If the screen only accept 8bit as reported by nvidia control panel, it may very well be that the screen is actually not a 8bit+FRC but a 8bit one.

[huhn]

Correct.
If the screen only accept 8bit as reported by nvidia control panel, it may very well be that the screen is actually not a 8bit+FRC but a 8bit one.

he is using a 120 hz screen could be a limitation from DP 1.2 this maybe works with DVI.
but the eizo is know to be 8 bit FRC

[nevcairiel]

Correct.
If the screen only accept 8bit as reported by nvidia control panel, it may very well be that the screen is actually not a 8bit+FRC but a 8bit one.

Screens also do some color processing, so it might be 8+FRC and still not accept 10-bit input. Its a gaming screen afterall..

[Shiandow]

Using 10bit setting in madVR is a GREAT idea for 8bit + FRC monitors.
For the simple reason that FRC dithers in the refresh rate of the display, but madVR is tied to the frame rate of the video played (until of course madshi decides to change that, if ever).
And because madVR does not dither the perfect integers of 10bit output but 8bit+FRC display is, so not double dithering, everything falls into place like magic.

I'm pretty sure MadVR dithers in the refresh rate of the display if you enable smooth motion. Also, if you output 10 bit then MadVR should dither from 16 bits to 10, which is then dithered from 10 bits to 8 on your monitor, this would be worse than directly dithering to 8 bits. Although it's unlikely that you'll be able to actually see any difference.

[Anime Viewer]

Does watching movies in 10bit will be any usefulness ? what will it be good for if not for movies

The usefulness of you running in 10-bit depends on a number of factors. 10-bit provides access to potentially more colors to be displayed. If you don't notice more colors, or an increase in the different shades of colors in the videos you are watching then it will not be useful to you. If your source is providing 8-bit instead of 10-bit the visual you are given will be inferior to what you would have seen if you had a 10-bit source being displayed at 10-bit. If you're screen only supports up to 8-bit then again you will not see the benefit provided by 10-bit.

If you're not noticing a difference between when you run in 8-bit mode compared to 10-bit mode, or you like taking screen shots of full screen video (which doesn't work in exclusive mode) then you might as well stick with the 8-bit setting(s).

[James Freeman]

I'm pretty sure MadVR dithers in the refresh rate of the display if you enable smooth motion. Also, if you output 10 bit then MadVR should dither from 16 bits to 10, which is then dithered from 10 bits to 8 on your monitor, this would be worse than directly dithering to 8 bits. Although it's unlikely that you'll be able to actually see any difference.

Good to know about smooth motion dithering at refresh speed, thanks.
So when smooth motion is enabled the dithering by madVR can be considered of better quality than FRC.

Isn't madVR leaves the steps that don't need dithering undithered?
Can we say the same for FRC?

[Ver Greeneyes]

Technically the FRC dithering could be happening at a higher rate than the overall refresh rate since it just has to toggle between two adjacent colors. One disadvantage of dithering on the monitor is it probably doesn't take into account nonlinearities in the display - e.g. for 10.5 it'll probably just alternate between 10 and 11, even if the correct ratio would be something like 25% to 75% (which madVR could account for with its linear light scaling). Admittedly the difference is likely to be small though.

[huhn]

Good to know about smooth motion dithering at refresh speed, thanks.
So when smooth motion is enabled the dithering by madVR can be considered of better quality than FRC.

Isn't madVR leaves the steps that don't need dithering undithered?
Can we say the same for FRC?

it doesn't dither at refreshrate speed it dithers at output speed that could 48 FPS at 60 hz with a 24p source.

[James Freeman]

Technically the FRC dithering could be happening at a higher rate than the overall refresh rate since it just has to toggle between two adjacent colors.

Reading this article: http://robotics.ee.uwa.edu.au/eyejr/lcd/FRC-information.pdf


Frame Rate Control is achieved by tuning pixels on and off over several frame
periods. With sufficient frame refreshing frequency, our human eyes will average
out the darkness of a pixel so that the individual pixel will show as gray.


Besides, sufficient high frequency is necessary in order to achieve a smooth display.

This works by combining four colour frames as a sequence in time, resulting in perceived mixture.

Man, I sure hope that you are right and FRC uses at least four times the refresh rate of the screen to generate the shades.
If not, and FRC uses the refresh rate of the screen, it is the worst method by far and should be used strictly for static images.

[tickled_pink]

Regarding Debanding - I can't say one is better than the other. Just when I think one is better, I switch to a different source to see that my preference has changed. My personal preference is "less is more" so I generally prefer the lowest setting. I'd rather see a bit of banding than a washed out scene. My "vote" goes to the more efficient algo which appears to be MadVR (according to GPU-Z).

[cvrkuth]

I use PotPlayer 1.6.54133 64bit with madVR v0.88.8 (64-bit) on Windows 8.1 (64-bit).
I noticed that when I press pause the video freezes. When press pause again, the video is still frozen, but the sound is continue playing.

This occurs only with 64-bit madVR D3D11 FSE mode 10bit (D3D9 FSE 8bit works well)

[Barnahadnagy]

After some tests in the previous days, I came to the conclusion that I have no conclusive opinion. MadVR's algo does a fine job at most cases, then there are some where Shiandow's is better. The same could be said for detail preservation, thou MadVR's high felt a bit worse overall in this regard. Overall, Shiandow's algo matches MadVR's in its latest iteration (At the beginning I preferred MadVR's, but it improved a lot. Good job, Shiandow!). If anything, the "High" preset should be tested against Shiandow's more, but I'm afraid this is all I could do.

[vivan]

Man, I sure hope that you are right and FRC uses at least four times the refresh rate of the screen to generate the shades.
If not, and FRC uses the refresh rate of the screen, it is the worst method by far and should be used strictly for static images.I have 6-bit + FRC IPS display, dithering noise is easy noticeable, so I'm pretty sure that it runs at 60 hz. But I can't spot the pattern. When comparing it with true 8-bit IPS (and madVR set to 6-bit) - I would say that dithering is far from dynamic ordered - more like dynamic random.
And banding there is terrible, 6-bit output in madVR is the only way to fight it.

But this might be true for those 200+ inflated hz TV panels.

[James Freeman]

I have 6-bit + FRC IPS display, dithering noise is easy noticeable, so I'm pretty sure that it runs at 60 hz. But I can't spot the pattern. When comparing it with true 8-bit IPS (and madVR set to 6-bit) - I would say that dithering is far from dynamic ordered - more like dynamic random.
And banding there is terrible, 6-bit output in madVR is the only way to fight it.

Great, thanks for that vivian.

I found a great read about FRC that clearly explains what it is: http://m.prohardver.hu/dl/cnt/2007-03/1500/iccz/hi-frc.pdf
As I understand Spatial dithering is a group of static pixels in a pattern, FRC is Temporal dithering is averaged by the human eye because of fast change between shades.
Like you already noticed, FRC in displays is nothing more than Spatial and Temporal combined, or "Random Dithering" with "Change Every Frame" at the monitor refresh rate, in madVRs language.

When I test in 3bit I can clearly see that enabling Smooth Motion makes the dithering pattern faster, while when disabling SM the pattern changes every video frame (24).
But this is still not as fast as the refresh rate, because when I speed up the video the dithering gets even faster till it reaches max speed of the refresh rate.
So madVR is worse than FRC in that sense that it is not using the maximum potential of Temporal dithering.
I hope madshi will realize that it is better to dither every refresh of the display, instead it being somehow related to the slower video frame rate.


On the worse side of understanding, FRC (Temporal dithering) needs at least 4 refresh cycles to generate the proper shade of the promised higher bit depth in a single pixel.
Since 6 to 8bit, and 8 to 10bit, has exactly 4 times the steps,
For example:
For 6+FRC (60Hz) to appear as 8bit, the maximum video frame rate has to be 4 times slower (and slower) than the refresh rate - > 15fps max on a 60Hz display.
For a 8+FRC display to look like a true 10bit display same principle applies, the image has to stay still for at least 4 refresh cycles.

[Asmodian]

Man, I sure hope that you are right and FRC uses at least four times the refresh rate of the screen to generate the shades.
If not, and FRC uses the refresh rate of the screen, it is the worst method by far and should be used strictly for static images.

Because no IPS LCD has been released with real response times <5 ms, and almost all are >8ms for the slowest transitions, I am pretty sure all the temporal dithering happens at the refresh rate, not faster. Even the fastest TN panels have max transition times of ~3 ms, too slow for above refresh rate dithering.

I prefer error diffusion 1 (mono, static) and madVR set to 6-bit on a 6-bit+FRC panel. I don't know about 8-bit+FRC because none of my current displays accept >8-bit at all (the Nvidia control panel only lists 8-bit, using DP1.2). I find temporal dithering distracting, it adds a slight shimmer to the image.

[MistahBonzai]

Also keep windows in 10bit so that the switch to FSE is faster.
Using native 10-bit 4:4:4 display with Direct3D 11/FSE. Works great! But for the life of me I can't figure out how/where to set windowed mode to 10-bit. BTW: My GPU is AMD HD-7850 if that matters (like with windowed overlay) and running fully updated Windows 7. Help?

[YxP]

I'm very familiar with the madVR settings but for the life of me I can't figure out how/where to set windowed mode to 10-bit. BTW: My GPU is AMD HD-7850 if that matters (like with windowed overlay) and running fully updated Windows 7. Help?

10bit works only in fullscreen exclusive mode.

[Patrik G]

one question here
Do you ever watch the movies or are you looking forward to another 2 hours of pixel peeping the dithering ;)?

set up your settings and let it be.
i takes one hour max then enjoy the damn movies instead :)
like its ment to be used lol

[MistahBonzai]

10bit works only in fullscreen exclusive mode.

That's what I thought but James Freeman seems to indicate otherwise in the last line of post 30217 (http://forum.doom9.org/showpost.php?p=1722951&postcount=30217)

[nevcairiel]

That's what I thought but James Freeman seems to indicate otherwise in the last line of post 30217 (http://forum.doom9.org/showpost.php?p=1722951&postcount=30217)

In the latest NVIDIA drivers you can set the driver to always output 10/12-bit over DP/HDMI. This does not change what Windows does, and you still need FSE to output 10-bit with madVR - it does however make switches from/to FSE faster, as the driver does not have to change output format, so the DP/HDMI link does not have to be re-initiated.

[FireFreak111]

Because no IPS LCD has been released with real response times <5 ms, and almost all are >8ms for the slowest transitions, I am pretty sure all the temporal dithering happens at the refresh rate, not faster. Even the fastest TN panels have max transition times of ~3 ms, too slow for above refresh rate dithering.

I prefer error diffusion 1 (mono, static) and madVR set to 6-bit on a 6-bit+FRC panel. I don't know about 8-bit+FRC because none of my current displays accept >8-bit at all (the Nvidia control panel only lists 8-bit, using DP1.2). I find temporal dithering distracting, it adds a slight shimmer to the image.

Current best IPS panel in reference to response times:
http://www.tftcentral.co.uk/reviews/acer_xb270hu.htm
On the normal overdrive setting at 144hz there was practically no overshoot, and an average of 5.5ms real response time. There is no FRC on this display, as far as I can tell its an 8 bit panel, not 8 bit + FRC. Your point still stands.

Ideally you would need an OLED panel to get the response times your thinking of, but at that point you might as well make the panel 10 bit and cut the need for FRC, since they are still incredibly expensive.

[Asmodian]

Current best IPS panel in reference to response times:
http://www.tftcentral.co.uk/reviews/acer_xb270hu.htm
On the normal overdrive setting at 144hz there was practically no overshoot, and an average of 5.5ms real response time. There is no FRC on this display, as far as I can tell its an 8 bit panel, not 8 bit + FRC. Your point still stands.

Ideally you would need an OLED panel to get the response times your thinking of, but at that point you might as well make the panel 10 bit and cut the need for FRC, since they are still incredibly expensive.

That 144Hz IPS is actually my current display and it is why I had to say below 5ms. ;)

And yes, it only accepts 8-bit, but it is still a surprisingly good panel.

Oddly on a 8-bit+FRC monitor setting Windows to 10-bit (if possible) would be better but with madVR I think I would still prefer 8-bit output due to madVR's dithering. :)

[JarrettH]

Don't drive yourselves crazy :D

[James Freeman]

Oddly on a 8-bit+FRC monitor setting Windows to 10-bit (if possible) would be better but with madVR I think I would still prefer 8-bit output due to madVR's dithering. :)
144Hz IPS is actually my current display


8bit+FRC at 144Hz would be better than 8bit+FRC at 60Hz, for the simple reason that it is faster.
It takes 4 refresh cycles to simulate the 4 shades difference between 8 and 10bit in a single pixel, so higher refresh rate is significant.
For a reason unknown to me, madVR does not dither in the screens refresh rate thus not utilizing the true potential of temporal dithering.
Hence FRC is still better to some extent, although madVR dithering is gamma compensated but this becomes less significant the higher the bit depth.

Madshi can make madVR dithering better than any FRC just by making an option to dither at displays refresh rate.
When/If it is done, 8bit + madVR dithering would be better than any 8+FRC, especially at higher refresh panels.

[MistahBonzai]

In the latest NVIDIA drivers you can set the driver to always output 10/12-bit over DP/HDMI. This does not change what Windows does, and you still need FSE to output 10-bit with madVR - it does however make switches from/to FSE faster, as the driver does not have to change output format, so the DP/HDMI link does not have to be re-initiated.
Thanks Nev. I thought I had been overlooking some simple remedy for resolving the messy HDMI re-init between 10-bit FSE and 8-bit windowed mode with the AMD drivers.

[RainyDog]

In the latest NVIDIA drivers you can set the driver to always output 10/12-bit over DP/HDMI. This does not change what Windows does, and you still need FSE to output 10-bit with madVR - it does however make switches from/to FSE faster, as the driver does not have to change output format, so the DP/HDMI link does not have to be re-initiated.

Thanks for pointing this out nevcairiel. It didn't cross my mind even though the constant output format changes have been driving me nuts as my telly also switches Scene modes too. I'll update to the latest drivers tonight then.

I still don't understand why my custom resolution for perfect 23.976 playback is ignored when using D3D11 10bit output though. I get 23.971 instead leading to occasional frame skips. Unless the issue it that I don't need a custom resolution in 10bit mode? I'll try disabling it and see what happens.

[yok833]

Does someone know if the last generation of Panasonic plasma (st50/st60) can manage 10 bit ????if yes is it necessary to also change the nvidia control panel from 8 to 12 bit in addition of MadVR d3d11 & 10 bit output?

[phoenix369]

In the latest NVIDIA drivers you can set the driver to always output 10/12-bit over DP/HDMI. This does not change what Windows does, and you still need FSE to output 10-bit with madVR - it does however make switches from/to FSE faster, as the driver does not have to change output format, so the DP/HDMI link does not have to be re-initiated.

Could this be the most certain indicative of a monitor support for 10-bit?

[Shiro Baka Neko]

I'm pretty sure MadVR dithers in the refresh rate of the display if you enable smooth motion. Also, if you output 10 bit then MadVR should dither from 16 bits to 10, which is then dithered from 10 bits to 8 on your monitor, this would be worse than directly dithering to 8 bits. Although it's unlikely that you'll be able to actually see any difference.

Then what should I set for dithering for 6bit+FRC 60Hz monitor?
I think 6bit output and error diffusion 1 look best, but my gpu is too weak to run error diffusion.

[huhn]

8bit+FRC at 144Hz would be better than 8bit+FRC at 60Hz, for the simple reason that it is faster.
It takes 4 refresh cycles to simulate the 4 shades difference between 8 and 10bit in a single pixel, so higher refresh rate is significant.
For a reason unknown to me, madVR does not dither in the screens refresh rate thus not utilizing the true potential of temporal dithering.
Hence FRC is still better to some extent, although madVR dithering is gamma compensated but this becomes less significant the higher the bit depth.

Madshi can make madVR dithering better than any FRC just by making an option to dither at displays refresh rate.
When/If it is done, 8bit + madVR dithering would be better than any 8+FRC, especially at higher refresh panels.

you know you need 4 cycles to get 10 bit with 8 bit FRC. madVR needs 8 bit to get 11 bit.
my 144 hz 6 bit+FRC doesn't even come close to 8 bit unditehred and that with a still image.
because it sound good on paper doesn't mean it is good at all.
maybe this helps maybe not. but 6 bit+FRC panel look usually pretty bad.
Does someone know if the last generation of Panasonic plasma (st50/st60) can manage 10 bit ????if yes is it necessary to also change the nvidia control panel from 8 to 12 bit in addition of MadVR d3d11 & 10 bit output?
plasma a low bit dither panel. they will nearly for sure support 10/12 bit input but if they benefit not you have to find this out by your self.

Could this be the most certain indicative of a monitor support for 10-bit?
if your can send 10 bit it can handle 10 bit input like most screen doesn't mean they have a 10 bit panel what so ever.

Then what should I set for dithering for 6bit+FRC 60Hz monitor?
I think 6bit output and error diffusion 1 look best, but my gpu is too weak to run error diffusion.
if 6 bit error diffusion looks best i wouldn't be shocked if 6 bit ordered dithering would still be better.
i guess you have to try out.

[phoenix369]

if you can send 10 bit it can handle 10 bit input like most screen doesn't mean they have a 10 bit panel what so ever.

So, if I understand correctly, this setting only affects what the GPU sends, and it´s not aware of the display capabilities.
If this were true, I suppose everybody with a modern GPU could force sending 10-bit output to the display. I´m telling this because that setting is locked at 8bpc with my GF970, HDMI 2.0 to my Panasonix AX800.

[huhn]

So, if I understand correctly, this setting only affects what the GPU sends, and it´s not aware of the display capabilities.
If this were true, I suppose everybody with a modern GPU could force sending 10-bit output to the display. I´m telling this because that setting is locked at 8bpc with my GF970, HDMI 2.0 to my Panasonix AX800.


hdmi 2.0 can't do 10 bit RGB UHD @ 60 HZ.
should be possible with DP this connector can do 10 bit RGB UHD @60hz or the display can't handle 10 bit input.

[leeperry]

With a 10-bit output it seems as though the smoother error diffusion should be all that is necessary.
Fully agreed, The lack of ED@10bit is also one of the reasons why I don't care much about getting my box to support it when ED2@8bit looks so wunderbar to begin with :D

Do you ever watch the movies or are you looking forward to another 2 hours of pixel peeping the dithering ;)?

set up your settings and let it be.
i takes one hour max then enjoy the damn movies instead :)
like it's meant to be used lol
Yeah yeah, well you're in the thread of Team Nitpick buddy. Much like audio gear you dunno what you've been missing till you see/hear it. Many inexperienced ppl will tell you that all DAC's sound the same yadayada, till they hear TOTL gear....and if they still don't hear a difference then they're very lucky coz they'll have cash to invest somewhere else.

Try mVR for a while enabling all its best features then go back to EVR if you dare :p

[nevcairiel]

Fully agreed, The lack of ED@10bit is also one of the reasons why I don't care much about getting my box to support it when ED2@8bit looks so wunderbar to begin with :D

Even a "worse" dithering in 10-bit will reduce the noise floor significantly compared to any output in 8-bit, that is assuming your screen actually does processing and display in 10-bit, because if it does not, sending 10-bit values is questionable at best anyway, and should be judged *extremely* carefully.
I would argue that if an image looks perceptibly better in 8-bit with ED than in 10-bit with OD, then your display is not suited for proper 10-bit image processing - which isn't uncommon at all, only very few absolute high-end displays are fully 10-bit capable.

In any case, there is absolutely zero prove that Error Diffusion in 10-bit would do any good, all you two have is suppositions.
Something looking better in 8-bit@ED could be considered the same concept as some content looking better with a soft scaler, it masks some inherent flaws due to stronger dithering. Even 10-bit ED would not solve that, in fact, it may make it worse!

I'm in no way against such an option, since it sounds like something very simple to do, i'm just not certain your arguments and observations match the most logical and likely behavior, which is the display dithering to 8-bit from the 10-bit input.

PS:
In before there is a trade quality for performance option in the future and you arrive at the same conclusion (which makes more sense than any alternatives). :p

[mbordas]

Does someone know if the last generation of Panasonic plasma (st50/st60) can manage 10 bit ????if yes is it necessary to also change the nvidia control panel from 8 to 12 bit in addition of MadVR d3d11 & 10 bit output?

I don't remember who first suggested it, but I ran the entech monitor asset manager:

http://www.entechtaiwan.com/util/moninfo.shtm

and it showed this:

CE vendor specific data (VSDB)
Supports 48bpp........... No
Supports 36bpp........... Yes
Supports 30bpp........... Yes
Supports YCbCr 4:4:4..... Yes

which I assume means 10 and 12 bit supported.

[6233638]

Even a "worse" dithering in 10-bit will reduce the noise floor significantly compared to any output in 8-bit, that is assuming your screen actually does processing and display in 10-bit, because if it does not, sending 10-bit values is questionable at best anyway, and should be judged *extremely* carefully.
I would argue that if an image looks perceptibly better in 8-bit with ED than in 10-bit with OD, then your display is not suited for proper 10-bit image processing - which isn't uncommon at all, only very few absolute high-end displays are fully 10-bit capable.To be clear, there was one test out of maybe 10 different ones where, subjectively, it could be argued that 8-bit plus Error Diffusion vs 10-bit with Ordered Dithered looked slightly better in 8-bit.
But with the blanking that the screen does when switching between 8-bit and 10-bit, without controls to change the type of dither used without exiting FSE, and without the option of using error diffusion with a 10-bit output, it's difficult to do a proper comparison.

Every other test I did was clearly in favor of 10-bit and undithered there are clearly four times as many steps of gradation on my display.
I know that the panel used in my TV is natively 10-bit, and when comparing 8-bit and 10-bit the difference is clear.

It was just this one test where it seemed as though you could maybe prefer the 8-bit image over 10-bit, so I wondered if it might be due to the difference in dither being used. Or perhaps it's something else entirely, since it was only one out of several tests which had this result. (perhaps even the test at fault)

In any case, there is absolutely zero prove that Error Diffusion in 10-bit would do any good, all you two have is suppositions.
Something looking better in 8-bit@ED could be considered the same concept as some content looking better with a soft scaler, it masks some inherent flaws due to stronger dithering. Even 10-bit ED would not solve that, in fact, it may make it worse!

I'm in no way against such an option, since it sounds like something very simple to do, i'm just not certain your arguments and observations match the most logical and likely behavior, which is the display dithering to 8-bit from the 10-bit input.Agreed. Error Diffusion may be completely unnecessary at 10-bit, but I'd like to have the option for comparison than assume that it is unnecessary, or that it would certainly be an improvement.

If it were not for the fact that it raises the black level so that it is never zero, which is problematic for a local dimming set like mine (means the backlight zones never switch off) I might actually be leaning towards random dither - especially at 10-bit where the noise level should be considerably lower.

Error diffusion can leave areas of solid color or patterns in the image, while random dither applies noise across the whole image equally.
As you get to higher resolutions and bit-depths it may not matter so much, but at 1080p random dither seems smoother.

Error Diffusion (http://abload.de/img/error-diffusion12ph4.png)
Random Dither (http://abload.de/img/random-ditherlpqq7.png)

Now this is an exaggerated example - it is highly magnified and I set the output to 4-bit to make the differences more obvious.
But you can see that random dither - though noisy - produces a continuous tone, while error diffusion (1) still has noticeable banding due to dither patterns or areas which are a solid color.

As you increase the bit-depth the noise gets lower and lower, and I wonder if random dither may even be the best option at 10-bit on most displays - especially if we get the option to change the dither pattern every refresh rather than every frame.
Or we may find that as you get to 10-bit, this "banding" with error diffusion is no longer noticeable at all, and the lower noise level of error diffusion is preferable.

Of course it may also be that once you are at 10-bit on a true 10-bit display the dither pattern really does not matter at all since the differences between each shade are so small.


Does someone know if the last generation of Panasonic plasma (st50/st60) can manage 10 bit ????if yes is it necessary to also change the nvidia control panel from 8 to 12 bit in addition of MadVR d3d11 & 10 bit output?If the option to select a higher bit-depth is available in the NVIDIA Control Panel, it should mean that the display supports it.
madVR will switch the display into 10-bit mode when entering DX11 FSE mode if it is supported, so it is not necessary to set that in the control panel, however.

I do wonder whether the display would be able to take advantage of a 10-bit input though. Plasmas make use of heavy dithering to draw their image, as they are technically "1-bit" displays just like DLPs.
But it may still be beneficial to send the display a 10-bit signal rather than 8-bit for its internal processing.
I'd be inclined to send any display the highest bit-depth that it supports on an input, because it will almost certainly be applying additional processing regardless of what bit-depth the panel is, and a higher bit-depth input should result in less errors when doing so.

[huhn]

I don't remember who first suggested it, but I ran the entech monitor asset manager:

http://www.entechtaiwan.com/util/moninfo.shtm

and it showed this:

CE vendor specific data (VSDB)
Supports 48bpp........... No
Supports 36bpp........... Yes
Supports 30bpp........... Yes
Supports YCbCr 4:4:4..... Yes

which I assume means 10 and 12 bit supported.

now you learn you can input up to 12 bit doesn't say anything about the panel it self. just a look at the driver give you the same information.

[x7007]

now you learn you can input up to 12 bit doesn't say anything about the panel it self. just a look at the driver give you the same information.

Should I select Ordered Error Diffusion when I use 10 bit FSE ? or it automatically change to Ordered

http://i.imgur.com/xXhRZkW.png

[huhn]

Should I select Ordered Error Diffusion when I use 10 bit FSE ? or it automatically change to Ordered


10 bit doesn't use ED. only use ED if you see a clear improvement with it.

[Qaq]

I see some still think 8 bit output is fine for 8 bit (and less) panels. But what we have learned (I hope) from recent discussions is that madVR output is just intermediate result between madVR and display video processing. And since 8 bit is enough, why not ask for trade option "process in 8 bits" then? That should unload GPU a lot! Weak GPUs for weak panels. No EVR anymore, anywhere.
Just kidding, but though.

[pirlouy]

@Madshi: About the fix you wrote for Windows 8 59/60 23/24 Hz auto refresh rate change. It works well for my ATI HD 5770, but with Intel HD 2000, it can't prevent a quick change in refresh rate. I mean

Without your fix, with Windows 8 bug
59 Hz -> no changes
60 Hz -> 59 Hz

With your fix on ATI 5770
59 Hz -> no changes
60 Hz -> no changes

With your fix on Intel HD 2000
59 Hz -> no changes
60 Hz -> quick change to 59 Hz -> quick changes to 60 Hz, each time I enter FullScreen (exclusive)

Let's hope it will be fixed by Microsoft for Windows 10.

[Ver Greeneyes]

I see some still think 8 bit output is fine for 8 bit (and less) panels. But what we have learned (I hope) from recent discussions is that madVR output is just intermediate result between madVR and display video processing. And since 8 bit is enough, why not ask for trade option "process in 8 bits" then? That should unload GPU a lot! Weak GPUs for weak panels. No EVR anymore, anywhere.
Just kidding, but though.GPUs are best at manipulating floating point numbers, and there's already the "store custom pixel shader results in 16bit buffer instead of 32bit" trade quality option which presumably makes it use 16-bit floats (a variant of the 'half' type in HLSL terms). 8-bit integers aren't even available in HLSL, though I guess not every part of madVR uses that. So aside from visibly losing information on the least significant bits, it probably wouldn't even help performance.

[leeperry]

In before there is a trade quality for performance option in the future and you arrive at the same conclusion (which makes more sense than any alternatives).
Options are good that's all, as all displays respond differently and my red deficient eyes got hooked to ED ever since the SmoothAdjust() Avisynth plugin half a decade ago. Anyway, PQ is stellar atm but I'll keep the 10bit option on the back burner.