madVR - high quality video renderer (GPU assisted)

[AngelGraves13]

Ok, so after some testing, I've settled on NGU-low for Chroma and Doubling (always if needed) and Soften Edges 1 and Add Grain 4. The grain really helps hide compression artifacts.

I've disabled SR and Remove Ringing Artifacts as it seems to degrade the image with NGU.

[Asmodian]

Originally Posted by Crimson Wolf

[*]image downscaling: catmull AR or bicubic 150 (same thing)

catmull AR is not the same as bicubic 150, catmull is bicubic 50.

[Crimson Wolf]

Originally Posted by Asmodian

catmull AR is not the same as bicubic 150, catmull is bicubic 50.

Ah, I'm sorry. I've switched to using SSIM for downscaling a while back. I must have got that part wrong.

@Growdelan

I just noticed your display is 1680x1050, so not a 1.5x scaling from original 720 video. That along with soften edges might be reason NGU is looking less sharp. On my display, when I do 720p > 1080p scaling, no matter which image upscaling and upscaling refinement settings I try, it would look worse than luma double and downscaling with SSIM, so I switched.

[Growdelan]

Originally Posted by Crimson Wolf

@Growdelan

I just noticed your display is 1680x1050, so not a 1.5x scaling from original 720 video. That along with soften edges might be reason NGU is looking less sharp. On my display, when I do 720p > 1080p scaling, no matter which image upscaling and upscaling refinement settings I try, it would look worse than luma double and downscaling with SSIM, so I switched.

I tried on the TV, but the result the same

http://screenshotcomparison.com/comparison/191841

[Backflash]

Originally Posted by Growdelan

I tried on the TV, but the result the same

http://screenshotcomparison.com/comparison/191841

I'm pretty sure that's what Bicubic 150 does.
Try SSIM 1D 100% AR/LL(optional)/AB 50 or 100% depending on your taste. I think 660 can handle it.

[huhn]

super XBR is sharper in this case but NGU has less ringing.

[Growdelan]

Originally Posted by Backflash

I'm pretty sure that's what Bicubic 150 does.
Try SSIM 1D 100% AR/LL(optional)/AB 50 or 100% depending on your taste. I think 660 can handle it.

SSIM 1D100% + AR, LL, AB 100% - did not help, lips and eyes sharper on super-xbr + lanczos3.

http://screenshotcomparison.com/comparison/191909

[aufkrawall]

I never understood why there is AB for SSIM.

[Uoppi]

Originally Posted by Crimson Wolf

In the end, trust in your eyes. If you can't tell the difference, does it matter?

Maybe it shoudn't but...

Looking at various zoomed-in screenshot comparisons, it's more or less easy to see some algorithms are simply better (in terms of artefacts etc.)

Even if you couldn't see those differences at all under normal viewing conditions (like I don't think I can for chroma, for example), just knowing you're using the "worst" settings may degrade the experience. If your GPU can handle it, why not use a decent algorithm "just in case"?

Because placebo is a very real and powerful phenomenon*, I see no reason why using "recommended best" settings wouldn't be a valid and acceptable way to proceed.

(* I've personally heard noteworthy differences between regular and high-end audio cables - only to learn the tester actually did not switch any cables at all... )

[hu1kamania]

Enjoying NGU for 720p to 1080p, great job Madshi!

I'm sticking with NNEDI3 for lower quality sources. NGU seems to bring out compression artifacts with it's extra detail.

[Backflash]

Originally Posted by Growdelan

SSIM 1D100% + AR, LL, AB 100% - did not help, lips and eyes sharper on super-xbr + lanczos3.

http://screenshotcomparison.com/comparison/191909

Disable soften edges, add grain 2.

[Ver Greeneyes]

Originally Posted by Uoppi

Even if you couldn't see those differences at all under normal viewing conditions (like I don't think I can for chroma, for example), just knowing you're using the "worst" settings may degrade the experience. If your GPU can handle it, why not use a decent algorithm "just in case"?

Because placebo is a very real and powerful phenomenon*, I see no reason why using "recommended best" settings wouldn't be a valid and acceptable way to proceed.

In addition, what if you test each algorithm individually and can't see a difference from a distance, but it turns out that adding them all together is just enough to make the difference visible? I'd rather be on the safe side and assume that it might make a difference even if I have to get up close to make out differences between individual algorithms. Besides, even if you can't consciously put your finger on what the difference is, a clearer image might reduce eyestrain or something.

[Uoppi]

Originally Posted by Ver Greeneyes

In addition, what if you test each algorithm individually and can't see a difference from a distance, but it turns out that adding them all together is just enough to make the difference visible? I'd rather be on the safe side and assume that it might make a difference even if I have to get up close to make out differences between individual algorithms. Besides, even if you can't consciously put your finger on what the difference is, a clearer image might reduce eyestrain or something.

Very good point. I can immediately see madvr improves the overall PQ (less aliasing and ringing being the most obvious), so all is certainy not placebo and I wouldn't pay for a dedicated GPU just for placebo anyway.

My "problem" atm is it's difficult to find sufficient time for testing when you're required to be changing diapers at the same time, lol. So any starting point recommendations are always welcome and this forum has been invaluable.

[ryrynz]

Originally Posted by Growdelan

SSIM 1D100% + AR, LL, AB 100% - did not help, lips and eyes sharper on super-xbr + lanczos3.

I'd try disabling AB for starters and then as above, disable SE if it's enabled.

[Growdelan]

Originally Posted by Backflash

Disable soften edges, add grain 2.

Originally Posted by ryrynz

I'd try disabling AB for starters and then as above, disable SE if it's enabled.

Ok, I turned off SE, and turned on AG 2. I think it work

http://screenshotcomparison.com/comparison/191961

[flossy_cake]

Any chance to get g-sync working in windowed mode? Gsync officially supports windowed mode applications so it should work with MadVR too.



It is really disappointing to spend $1k on a monitor and not be able to take advantage of this feature.

Note: On my system, Gsync does work in full screen exclusive mode, but only with DX11 disabled under Rendering - General Settings. But it doesn't work for all video files. Strangely it does work for all youtube videos I've tested so far (piped by SVPTube).

[huhn]

g-sync is a banned topic here.

madVR aims at HTPCs not gaming devices.

[flossy_cake]

MadVR aims at the best possible rendering performance. If you look at the extent Madshi has gone to increasing the picture quality, how many countless hours must have been spent pursuing image fidelity through all of the features implementation that currently exist, it's a no brainer that gsync compatibility would be a desirable feature to support. Odd you would consider refresh rate matching some kind of "gaming" gimmick. No different to using 24hz mode on a HDTV or the refresh rate matching listview in MPC-HC. The goal is the same, the implementation is just more intuitive with gsync because it doesn't require anything extra than rendering in a particular D3D mode.

[huhn]

http://forum.doom9.org/showpost.php?...ostcount=29243

[CruNcher]

Originally Posted by flossy_cake

Any chance to get g-sync working in windowed mode? Gsync officially supports windowed mode applications so it should work with MadVR too.



It is really disappointing to spend $1k on a monitor and not be able to take advantage of this feature.

Note: On my system, Gsync does work in full screen exclusive mode, but only with DX11 disabled under Rendering - General Settings. But it doesn't work for all video files. Strangely it does work for all youtube videos I've tested so far (piped by SVPTube).

So only because your ego spent 1K on it you think Markets will drift for you, that's not how they work

Also the whole stuff is so much in Research State at all and so much in flux that it's a pain to make decisions based on that, that tomorrow could go invalid.

I can understand Madshi very good not playing this "Game" *g* a rather sane decision on a developer resource side.