Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Hardware & Software > Software players

Reply
 
Thread Tools Search this Thread Display Modes
Old 16th February 2014, 01:06   #23241  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
Humm, that's plain ridiculous. The AGL1 build looks so much better than NL6.....need to test again tomorrow morning but the latter looks unnecessarily grainy and unfocused compared to the AGL1 build. The LL builds still look grainy and plain ugly, I don't see how that could change with fresh eyes but we'll see
leeperry is offline   Reply With Quote
Old 16th February 2014, 01:43   #23242  |  Link
Shiandow
Registered User
 
Join Date: Dec 2013
Posts: 753
Well, I suppose I should call myself a "scientist" then. Anyway based on the images 6233638 posted I'd say that adaptive is better. In an effort to make this somewhat more objective I also tried blurring the images. This confirmed that adaptive gave the 'smoothest' gradient and that unlimited was both more noisy and too dark.

Edit: I even think a very low level of this kind of noise might be appropriate for a low-noise algorithm but let's first try and get the medium noise algorithm right.

Last edited by Shiandow; 16th February 2014 at 01:46.
Shiandow is offline   Reply With Quote
Old 16th February 2014, 02:14   #23243  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 447
madshi, I was playing around with making my own ED algorithm to get a better understanding of it, and I was wondering: when you apply the limiter, do you update the error for that pixel to reflect that or do you just diffuse the error as if it wasn't limited?
Ver Greeneyes is offline   Reply With Quote
Old 16th February 2014, 02:58   #23244  |  Link
Shiandow
Registered User
 
Join Date: Dec 2013
Posts: 753
Quote:
Originally Posted by Ver Greeneyes View Post
madshi, I was playing around with making my own ED algorithm to get a better understanding of it, and I was wondering: when you apply the limiter, do you update the error for that pixel to reflect that or do you just diffuse the error as if it wasn't limited?
From the code he posted I'd say he did the former i.e. update the error by using the 'limited' pixel value. This seems to cause problems when limiting the pixel value makes it hard for a particular error to be distributed (see this post), madshi solved this by forcing the error to be smaller than the difference between adjacent colour values.

Last edited by Shiandow; 16th February 2014 at 03:11. Reason: Clarity
Shiandow is offline   Reply With Quote
Old 16th February 2014, 03:16   #23245  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 447
Quote:
Originally Posted by Shiandow View Post
From the code he posted I'd say he did the former i.e. update the error by using the 'limited' pixel value. This seems to cause problems when limiting the pixel value makes it hard for a particular error to be distributed (see this post), madshi solved this by forcing the error to be smaller than the difference between adjacent colour values.
Uh, right. I even looked at that code before. I guess I thought there would be more to it than that XD
Ver Greeneyes is offline   Reply With Quote
Old 16th February 2014, 06:14   #23246  |  Link
AngelGraves13
Registered User
 
Join Date: Dec 2010
Posts: 255
Does madvr support 64-bit MPC BE or HC yet?
AngelGraves13 is offline   Reply With Quote
Old 16th February 2014, 06:33   #23247  |  Link
rahzel
Registered User
 
Join Date: Jul 2005
Posts: 359
Posted this in the MPC thread because it seems to be more of an MPC issue, but hopefully someone can help me.

Anyone else having issues with MPC HC 1.7.2+ and Fullscreen Exclusive mode in MadVR? When I try to enter FSE mode on my Radeon 7750 HTPC, the bottom MPC bar stays at the bottom and it doesn't enter FSE mode. I have to exit and fullscreen it again. On my Intel system (HD4000), it switches to FSE mode every time, but I can see the bottom MPC bar stay there for a second or two, then it enters FSE. With MPC 1.7.1 and below, I have no issues at all. This also only happens with videos that have a ~1.85/1.78 AR... movies that are ~2.35+ (or have significant bars) enter FSE mode fine.
rahzel is offline   Reply With Quote
Old 16th February 2014, 07:14   #23248  |  Link
cyberbeing
Broadband Junkie
 
Join Date: Oct 2005
Posts: 1,859
Quote:
Originally Posted by madshi View Post
That wouldn't work because converting back to gamma corrected light gives me floating point values again.
Yes, I realize it would introduce error and not make much sense for dithering if you were then required to introduce more error afterwords, I just highly suspect doing that would resolve the gamma errors near-black & near-white.

Quote:
Originally Posted by madshi View Post
Only the error diffusion weight distributions and decisions are performed in linear light.
If you somehow gamma corrected the dither pixel results of the weight distributions and decisions after initially calculating them in linear light, it would achieve the same result (as gamma->linear->dither linear->gamma) while leaving the image in gamma corrected light the entire time. I don't have the slimmest idea how you would do this, though it must be possible considering dithering alters the RGB pixel values of an image. It just seems that if you did not desire linear light dithering to introduce gamma error on the near-black & near-white shade of the image, you would need to do something like this.

Quote:
Originally Posted by madshi View Post
Of course you can cover the whole gradient with loads of additional noise (like the "unlimited" build does) to make it look smoother in those contrast boosted screenshots. But that's really a "cover up", IMHO. The contrast is boosted in those screenshots so much that it can't really look smooth. But if you allow the different shades to be 8bit steps, the result should be perfectly smooth without any visible banding.
This isn't necessarily the case. As far as I'm aware, the HVS is more sensitive to slight changes in texture than it is color and luminance. Even if the image is technically perfectly smooth, it won't appear so unless the texture (or lack of texture) is perceptually even across all shades. If the dithering algorithm is the proper place for this to occur, who knows, since as you mentioned:
Quote:
Originally Posted by madshi View Post
you can probably achieve the same effect by simply adding noise/grain in a custom pixel shader

Last edited by cyberbeing; 16th February 2014 at 08:15.
cyberbeing is offline   Reply With Quote
Old 16th February 2014, 07:45   #23249  |  Link
AngelGraves13
Registered User
 
Join Date: Dec 2010
Posts: 255
FSE mode throws off sync for me, so I disabled it. The lips don't match 100% with the audio, it's slightly off. Enough to be annoying. Turning FSE mode off fixes the issue, so I've abandoned it until it's addressed.

Also...I don't see how MPC 65-bit not working with madvr (because it's 32-bit only) is a MPC issue. I'm asking when madvr will have a 64-bit renderer so that it can be used. Get the difference?
AngelGraves13 is offline   Reply With Quote
Old 16th February 2014, 07:50   #23250  |  Link
huhn
Registered User
 
Join Date: Oct 2012
Posts: 7,903
Quote:
Originally Posted by AngelGraves13 View Post
FSE mode throws off sync for me, so I disabled it. The lips don't match 100% with the audio, it's slightly off. Enough to be annoying. Turning FSE mode off fixes the issue, so I've abandoned it until it's addressed.

Also...I don't see how MPC 65-bit not working with madvr (because it's 32-bit only) is a MPC issue. I'm asking when madvr will have a 64-bit renderer so that it can be used. Get the difference?
maybe with 1.0 but there is no real benefit to 64 bit anyway...

just use 32 bit mpc hc nothing wrong with
huhn is offline   Reply With Quote
Old 16th February 2014, 07:51   #23251  |  Link
6233638
Registered User
 
Join Date: Apr 2009
Posts: 1,019
Quote:
Originally Posted by AngelGraves13 View Post
FSE mode throws off sync for me, so I disabled it. The lips don't match 100% with the audio, it's slightly off. Enough to be annoying. Turning FSE mode off fixes the issue, so I've abandoned it until it's addressed.
Try the "delay playback start" rendering options.
6233638 is offline   Reply With Quote
Old 16th February 2014, 09:00   #23252  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
6233638, Ver Greeneyes, bacondither

How do you create this effect to clearly see the dots (dithering), so I can see it first hand too.
What soft?

1. Take a screen shot.
2. ...
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.
James Freeman is offline   Reply With Quote
Old 16th February 2014, 09:15   #23253  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by James Freeman View Post
6233638, Ver Greeneyes, bacondither

How do you create this effect to clearly see the dots (dithering), so I can see it first hand too.
What soft?

1. Take a screen shot.
2. ...
1) You load the untouched images or the clips into madVR, skip to the same frame (if itīs a clip), go into windowed fullscreen, take a screenshot, use an image viewer/editor that is able to alter the input level range, adjust the max. level (should be at 255 originally) to something that shows you all bars or even lower to make lower steps more visible and save the image. I use GIMP for this (itīs free).

2) You basically do 1) with every new build, then you can compare (I use Irfanview for this, because it has a nice slideshow function that will make comparisons very easy) the end result.
iSunrise is offline   Reply With Quote
Old 16th February 2014, 09:28   #23254  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
Thanks iSunrise.

Yes I tweak the Input/Output levels in Paint.NET (also free and very compatible).
But still, I can't get that single bit dotted image like you guys post to compare the dithering versions.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.

Last edited by James Freeman; 16th February 2014 at 09:53.
James Freeman is offline   Reply With Quote
Old 16th February 2014, 10:49   #23255  |  Link
ryrynz
Registered User
 
ryrynz's Avatar
 
Join Date: Mar 2009
Posts: 3,646
Quote:
Originally Posted by madshi View Post
And NNEDI3 is another class, IMHO.
http://screenshotcomparison.com/comparison/59698
If only it didn't ruin the color and size of the x's and what's with all those vertical artifacts around the x's the numbers and letters?
I would love to see this somehow addressed, more neurons? (maybe not) or slightly better detection of things that should be left alone, even if it meant a loss of sharpness..

I don't know if NNEDI will see further development, It does so much so well, and fails hard in the rest.. I wish there was a way to minimize the problem areas, which I was hoping resizing before applying NNEDI might achieve
but that seems to kill much of the benefits of NNEDI.
ryrynz is offline   Reply With Quote
Old 16th February 2014, 10:59   #23256  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by ryrynz View Post
If only it didn't ruin the color and size of the x's and what's with all those vertical artifacts around the x's the numbers and letters?
I would love to see this somehow addressed, more neurons? (maybe not) or slightly better detection of things that should be left alone, even if it meant a loss of sharpness..

I don't know if NNEDI will see further development, It does so much so well, and fails hard in the rest.. I wish there was a way to minimize the problem areas, which I was hoping resizing before applying NNEDI might achieve
but that seems to kill much of the benefits of NNEDI.
I already mentioned this in a post of mine a couple of pages ago, along with an image of certain problem areas that look like ringing artifacts to me. An additional pass through madshiīs AR filter could probably eliminate almost all of them (if these are indeed added through NNEDI3 and arenīt in the source itself), but Iīm not sure if that would introduce other artifacts. Since thereīs no way for us to test that, I just wanted it to get out in the open as a thought.

Quote:
Originally Posted by James Freeman View Post
Thanks iSunrise.

Yes I tweak the Input/Output levels in Paint.NET (also free and very compatible).
But still, I can't get that single bit dotted image like you guys post to compare the dithering versions.
Here is the source (expanded original) that everyone seems to be using. Just do what I described above on it and it will work, it does for me.

Last edited by iSunrise; 16th February 2014 at 11:37.
iSunrise is offline   Reply With Quote
Old 16th February 2014, 11:32   #23257  |  Link
cyberbeing
Broadband Junkie
 
Join Date: Oct 2005
Posts: 1,859
Took a quick look at the latest builds.

Adaptive2 had the least banding on near-black and near-white gradients.
NL6-Gamma had the least banding on mid-tone gradients.
cyberbeing is offline   Reply With Quote
Old 16th February 2014, 13:56   #23258  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by AngelGraves13 View Post
Does madvr support 64-bit MPC BE or HC yet?
No.

Quote:
Originally Posted by AngelGraves13 View Post
FSE mode throws off sync for me, so I disabled it. The lips don't match 100% with the audio, it's slightly off.
That shouldn't happen. Seems most other users don't have this problem, so it's probably specific to your hardware or configuration. You could try lowering the size of the GPU queue and/or the number of frames that are presented in advance. Some GPU drivers don't like it if these numbers are too high.

Quote:
Originally Posted by leeperry View Post
Humm, that's plain ridiculous. The AGL1 build looks so much better than NL6.....need to test again tomorrow morning but the latter looks unnecessarily grainy and unfocused compared to the AGL1 build. The LL builds still look grainy and plain ugly, I don't see how that could change with fresh eyes but we'll see
Quote:
Originally Posted by Shiandow View Post
Well, I suppose I should call myself a "scientist" then. Anyway based on the images 6233638 posted I'd say that adaptive is better. In an effort to make this somewhat more objective I also tried blurring the images. This confirmed that adaptive gave the 'smoothest' gradient and that unlimited was both more noisy and too dark.
Ok, thanks for the feedback.

Quote:
Originally Posted by cyberbeing View Post
Took a quick look at the latest builds.

Adaptive2 had the least banding on near-black and near-white gradients.
NL6-Gamma had the least banding on mid-tone gradients.
And how does the first adaptive build fit into this? Which of three do you prefer?
madshi is offline   Reply With Quote
Old 16th February 2014, 14:09   #23259  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Just FYI, for anyone that has problems getting still images to show in madVR (black screen), you absolutely need to enable "delay playback start until render queue is full", otherwise madVR seems to have a problem displaying certain images (in my case, several 16bit/8bit per component TIFF and PNG files). Iīm not sure why it works for some, but not for others. Ver Greeneyes' PNG samples (posted the link above) show the problem, while other files work fine. I donīt quite get it.

madshi, is that a bug? If it is, Iīm going to add it to the bug tracker right away.

Last edited by iSunrise; 16th February 2014 at 14:12.
iSunrise is offline   Reply With Quote
Old 16th February 2014, 14:27   #23260  |  Link
Ziron
Registered User
 
Join Date: Nov 2013
Location: Canada
Posts: 2
Quote:
Originally Posted by ryrynz View Post
If only it didn't ruin the color and size of the x's and what's with all those vertical artifacts around the x's the numbers and letters?
I would love to see this somehow addressed, more neurons? (maybe not) or slightly better detection of things that should be left alone, even if it meant a loss of sharpness..

I don't know if NNEDI will see further development, It does so much so well, and fails hard in the rest.. I wish there was a way to minimize the problem areas, which I was hoping resizing before applying NNEDI might achieve
but that seems to kill much of the benefits of NNEDI.
I really wish tritical had open sourced his training suite for NNEDI. It would be interesting to see if training it on more/different data would produce less/different artifacts. Given the renewed interest in NNEDI recently though we might get lucky and have someone reverse engineer some new neural network training software for it.
Ziron is offline   Reply With Quote
Reply

Tags
direct compute, dithering, error diffusion, madvr, ngu, nnedi3, quality, renderer, scaling, uhd upscaling, upsampling

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 05:34.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.