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 17th February 2014, 20:14   #23361  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 447
Quote:
Originally Posted by iSunrise View Post
My understanding is that madshi implemented the "donīt use dithering" option in the "trade quality for performance" tab, because otherwise, there would always be dithering, even if you wouldnīt need upsampling/scaling at all. Is that a wrong assumption from my side? Otherwise why would he implement the "donīt use dithering" option? Only for lower bit-depth sources?

Also, it would require to always select "8bit (or higher)" under the "devices - properties" tab.
I believe what madshi implemented was to only do random dithering on pixels that aren't an integer multiple of 1/255. But with error diffusion I don't think that's relevant, since such pixels wouldn't have any error to diffuse (and if they're diffusing error from another pixel then you can't disable dithering for them anyway).

"don't use dithering", on the other hand, completely disables dithering, regardless of what you have your display's bit depth set to. Note that other things can still introduce dithering though, such as the decoder.
Ver Greeneyes is offline   Reply With Quote
Old 17th February 2014, 20:24   #23362  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by Ver Greeneyes View Post
I believe what madshi implemented was to only do random dithering on pixels that aren't an integer multiple of 1/255. But with error diffusion I don't think that's relevant, since such pixels wouldn't have any error to diffuse (and if they're diffusing error from another pixel then you can't disable dithering for them anyway).

"don't use dithering", on the other hand, completely disables dithering, regardless of what you have your display's bit depth set to. Note that other things can still introduce dithering though, such as the decoder.
I just saw that he originally spoke about 8-bit yuv444p and then he mentioned only "True 8-bit video should not be dithered at all, right?" in the second sentence, that confused the hell out of me for some reason. Should have read this more clearly.

Yes, the decoder may be another reason for dithering, however, LAV will pass-through the decoded result untouched (same that happens with the levels) to madVR if I remember correctly, so this thing is of the past for most people, which is actually another great accomplishment.

Last edited by iSunrise; 17th February 2014 at 20:42.
iSunrise is offline   Reply With Quote
Old 17th February 2014, 20:43   #23363  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 447
Quote:
Originally Posted by iSunrise View Post
I just saw that he originally spoke about 8-bit yuv444p and then he mentioned only "True 8-bit video should not be dithered at all, right?" in the second sentence, that confused the hell out of me for some reason. Should have read this more clearly.
Yeah, I made this mistake initially. YUV anything will still require conversion, but the 8-bit 'bgr0' encodes I made shouldn't.

Also a heads up to anyone using my test patterns: v2.1 is the same as before, but with the height reduced to 960 pixels to make getting a screenshot easier (now people with 1920x1080 monitors should be able to get a screenshot while running MPC-HC as a maximized window, for instance).
Ver Greeneyes is offline   Reply With Quote
Old 17th February 2014, 20:52   #23364  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
An Image(s) worth a thousand words:

Original (Untouched).

No Dithering (Low bitrate).

Random Dithering (Low bitrate dithered with MadVR Random Dithering).

ED Adaptive 4 (Low bitrate dithered with MadVR Adaptive 4).


Here's how:

1. In MPC-HC turn Brightness to -100, Contrast to +80 (its 16-bit through MadVR).
This to make the image dark.

2. In you favorite editor:
Input Level, to about 100 (from white 255) or leave it untouched.
Output Level Center, raise towards 255 till you think is enough (this is the important step).
Play with these till the image looks close to the Original.


I wish MadVR Brightness would go lower than -100, so that I would not need to raise that contrast too.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.

Last edited by James Freeman; 17th February 2014 at 21:14.
James Freeman is offline   Reply With Quote
Old 17th February 2014, 20:57   #23365  |  Link
The 8472
Registered User
 
Join Date: Jan 2014
Posts: 51
Quote:
Originally Posted by 6233638 View Post
Honestly, pixel art does not look good when you use video-style filters on it. It's just not how it was intended to look.
There are specialized pixel art scalers anyway: http://en.wikipedia.org/wiki/Image_scaling#Pixel_art_scaling_algorithms
The 8472 is offline   Reply With Quote
Old 17th February 2014, 21:11   #23366  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 447
Quote:
Originally Posted by James Freeman View Post
An Image(s) worth a thousand words: [...]
That's a pretty nice comparison
Ver Greeneyes is offline   Reply With Quote
Old 17th February 2014, 21:16   #23367  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,477
Quote:
Originally Posted by iSunrise View Post
Do you have a sample where you see this noise, because Iīm interested to compare it on my hardware-calibrated Eizo (that only has about 700:1 left after calibration to REC709).
Dunno, take a very clear HD movie such as Oblivion for instance, the sequence from 13'49 to 14'45 is one of my favorites when it comes to rolling ED's. Gosh, it looks so sharp and high contrast with ED4, too good

With NL6, it looks a hell more grainy and far less focused and even ED1 is a big downgrade when it comes to clarity and transparency.

I guess 700:1 is your bottleneck especially if there's a thick pearly anti-glare layer on your monitor. I'm always dubious when I see computer geeks raving about 1440p 24/27" monitors that use IPS panels and come with a very blurry anti-glare layer....they are not getting 1440p by a long shot IMO due to the blurriness.

Quote:
Originally Posted by iSunrise View Post
Now that we have A4, I can see a lot more noise in general. And it think thatīs a good thing, because that doesnīt seem to be the cause of the new dithering-algorithms at all, but it finally shows us a higher effective bit-depth of the original source, whereas the smearing/extreme noise of random dithering made that less visible.
Oh definitely, the A builds really get you to see the ugliness of your sources......64x NNEDI or else

Last edited by leeperry; 17th February 2014 at 21:20.
leeperry is offline   Reply With Quote
Old 17th February 2014, 21:28   #23368  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
Quote:
Originally Posted by Ver Greeneyes View Post
That's a pretty nice comparison
Thanks.

I think it is a nice method to test future builds.
AND, I gave my method for free !!!

madshi,
Any way to go lower than -100 on the Brightness?
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.
James Freeman is offline   Reply With Quote
Old 17th February 2014, 22:04   #23369  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by Ver Greeneyes View Post
That's a pretty nice comparison
Indeed, good job James Freeman.

When looking at this, I have at least two questions, though.

Since in James Freeman's original untouched screenshot, the black borders are as black as black gets (0) (I checked it), why is there dithering? Is that a rounding error?

Also, look at Bilboīs jacket around the high of his right hand, why is there no dithering applied at all (I also checked it, this is definitely not black)? There are several places where the really dark black areas donīt get any dithering applied to them. Is that an error of the algorithm?

Can someone else instead of madshi explains this to me?

Last edited by iSunrise; 17th February 2014 at 22:10.
iSunrise is offline   Reply With Quote
Old 17th February 2014, 22:07   #23370  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by iSunrise View Post
Since in James Freeman's original untouched screenshot, the black borders are as black as black gets (0), why is there dithering? Is that a rounding error?
My guess is that this is an artifact from the brightness/contrast modification inside madVR.

It would be better to do any enhancements *after* madVR, so that they do not influence the dithering behaviour.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is offline   Reply With Quote
Old 17th February 2014, 22:09   #23371  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by nevcairiel View Post
My guess is that this is an artifact from the brightness/contrast modification inside madVR.

It would be better to do any enhancements *after* madVR, so that they do not influence the dithering behaviour.
Yes, thatīs what I thought about, too. But why would it show like random dots? Doesnīt brightness or contrast increase/decrease everything (every pixel) by a certain level instead of just single pixels? Kinda odd or?
iSunrise is offline   Reply With Quote
Old 17th February 2014, 22:11   #23372  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by iSunrise View Post
Yes, thatīs what I thought about, too. But why would it show like random dots? Doesnīt brightness or contrast increase everything (every pixel) by a certain level instead of just single pixels? Kinda odd or?
Thats the error diffusion artifacts. It probably increased the value to something like 0.1 instead of 0, and the dithering accumulates the "error" and once it has enough, it'll draw a pixel with 1 instead of 0.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is offline   Reply With Quote
Old 17th February 2014, 22:13   #23373  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by nevcairiel View Post
Thats the error diffusion artifacts. It probably increased the value to something like 0.1 instead of 0, and the dithering accumulates the "error" and once it has enough, it'll draw a pixel with 1 instead of 0.
Ok, that makes sense, thanks nevcairiel.
iSunrise is offline   Reply With Quote
Old 17th February 2014, 22:13   #23374  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
Quote:
Originally Posted by nevcairiel View Post
My guess is that this is an artifact from the brightness/contrast modification inside madVR.
Correct.
No error of the algorithm.

Its because of the raised Contrast, it clips the darkest and brightest shades.

That's why I asked if there is more than -100 on Brightness.
So the all the information will still be there (16-bit).


EDIT:
The resulting non black bars are because of the elevated Contrast.
Even value of +1 or -1 creates colorful blacks.

Quote:
Originally Posted by nevcairiel View Post
It would be better to do any enhancements *after* madVR, so that they do not influence the dithering behaviour.
That's will not do, that will defy the purpose of this experiment.
Just the contrary, we need to use MadVR's Brightness because its 16-bit (to preserve all the information), then let it dither the resulting image.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.

Last edited by James Freeman; 17th February 2014 at 22:31.
James Freeman is offline   Reply With Quote
Old 17th February 2014, 22:13   #23375  |  Link
sneaker_ger
Registered User
 
Join Date: Dec 2002
Posts: 5,565
Quote:
Originally Posted by James Freeman View Post
An Image(s) worth a thousand words
Are the black bars encoded as 100% black?
sneaker_ger is offline   Reply With Quote
Old 17th February 2014, 22:21   #23376  |  Link
iSunrise
Registered User
 
Join Date: Dec 2008
Posts: 496
Quote:
Originally Posted by sneaker_ger View Post
Are the black bars encoded as 100% black?
Yes, at least when you look at the original png, they are.
iSunrise is offline   Reply With Quote
Old 17th February 2014, 22:25   #23377  |  Link
James Freeman
Registered User
 
Join Date: Sep 2013
Posts: 919
The answer to the non black borders is in my last post.
__________________
System: i7 3770K, GTX660, Win7 64bit, Panasonic ST60, Dell U2410.
James Freeman is offline   Reply With Quote
Old 17th February 2014, 22:37   #23378  |  Link
XMonarchY
Guest
 
Posts: n/a
Would enabling my i7 3770K HD4000 CPU graphics to work along with my GeForce GTX 770 improve madVR rendering performance? Would I need to get the Lucid Logix software to make it happen?

Also, LAV Video Decoder has can be set to use either random or ordered dithering. There is no way to disable it AFAIK. Does that mean that I also get dithering on top of error diffusion? Is that of benefit to me or not? If not, then is there a way to turn it off?

Finally, I set LAV Video to use nVidia CUVID acceleration for decoding. I know decoding is not the same as rendering, but since CUVID uses GPU, could it be reducing madVR rendering that also needs GPU power?
  Reply With Quote
Old 17th February 2014, 22:38   #23379  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by James Freeman View Post
Any way to go lower than -100 on the Brightness?
The brightness algorithm in madVR practically changes the gamma value, as simple as that. The way the current formula works, more than -100 is not possible, I'd have to change the formula for that. But you can further darken the image by enabling gamma processing and selecting a higher gamma value, e.g. 2.60.
madshi is offline   Reply With Quote
Old 17th February 2014, 22:38   #23380  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,344
Quote:
Originally Posted by XMonarchY View Post
Also, LAV Video Decoder has can be set to use either random or ordered dithering. There is no way to disable it AFAIK. Does that mean that I also get dithering on top of error diffusion? Is that of benefit to me or not? If not, then is there a way to turn it off?
Dithering in LAV Video is only used when it needs to be used, and if you use madVR and didn't disable any output formats in LAV, then it will never be needed.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel 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 22:37.


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