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 23rd July 2015, 13:56   #32081  |  Link
Thunderbolt8
Registered User
 
Join Date: Sep 2006
Posts: 2,171
Quote:
Originally Posted by Akeno View Post
If we do that, we get back into subjectivity.
why, in how far is judging this different than judging it from an image which has been up- and downscaled? how can you only be sure the result you see is from the algorythm you want to look at and has not been influenced by the other one?
__________________
Laptop Acer Aspire V3-772g: i7-4202MQ, 8GB Ram, NVIDIA GTX 760M (+ Intel HD 4600), Windows 8.1 x64, madVR (x64), MPC-HC (x64), LAV Filter (x64), XySubfilter (x64)
Thunderbolt8 is offline   Reply With Quote
Old 23rd July 2015, 14:21   #32082  |  Link
leeperry
Kid for Today
 
Join Date: Aug 2004
Posts: 3,462
Quote:
Originally Posted by Akeno View Post
I'm still under the impression that HQ is closer to the ground truth than LQ.
Fair enough, I'm sure madshi will explain me how wrong I am using 500% magnified close-ups but at the end of the day I kinda rest my case that SR is in dire need of at least 3 separate knobs for number of passes, strength and LQ/HQ.

Quote:
Originally Posted by aufkrawall View Post
256 neurons really improve the image a lot. However, 64 neurons are much more reasonable and still look better than super-xbr (e.g. regarding the antennas).
I meant to try 256 neurons NNEDI3 for chroma+luma but mVR .20 was refusing to scale the picture to 1080p for some reason, I seriously wish all this upscaling madness could be used on still pictures too coz it's either a black frames feast or a total refusal to show anything. All picture viewers provide craptastic upscaling algorithms.

Oh BTW I had a gamut mapping PS script engaged in PotP so yeah the ground truth is the one I labelled as is and not the untouched capture.

Last edited by leeperry; 24th July 2015 at 00:36.
leeperry is offline   Reply With Quote
Old 23rd July 2015, 14:49   #32083  |  Link
aufkrawall
Registered User
 
Join Date: Dec 2011
Posts: 1,708
Quote:
Originally Posted by leeperry View Post
I meant to try 256 neurons NNEDI3 for chroma+luma but mVR .20 was refusing to scale the picture to 1080p for some reason
Any idea why? I had doubling activated for >= factor 1.5 (default) and it worked just normally.

Btw: Is it technically possible to just sharpen chroma to fight blur of CSS?
aufkrawall is offline   Reply With Quote
Old 23rd July 2015, 14:53   #32084  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 445
Quote:
Originally Posted by leeperry View Post
I seriously wish all this upscaling madness could be used on still pictures too coz it's either a black frames feast or a total refusal to show anything.
I've had success loading images in MPC-HC, though it's not ideal for saving the result since I think madVR uses slightly different logic for that. I'm pretty sure the upscaling works just fine though.
Ver Greeneyes is offline   Reply With Quote
Old 23rd July 2015, 16:48   #32085  |  Link
digitech
Registered User
 
Join Date: Jun 2012
Posts: 54
Quote:
Originally Posted by nevcairiel View Post
Its a known issue, one that cannot be fixed properly (ie. without copy-back), which is why you should use copy-back.
madVR used to have its own copy-back function, but it was removed because LAV is just more efficient at it.

It seems to mostly affect NVIDIA, which is the good news, since copy-back is very efficient on nvidia anyway.
Which do you think is more efficcient? Im on nvidia 760 and still using Cuvid in Lav Filters, is copy back a better alternative?
digitech is offline   Reply With Quote
Old 23rd July 2015, 17:54   #32086  |  Link
aufkrawall
Registered User
 
Join Date: Dec 2011
Posts: 1,708
It should be in no way inferior. You probably even save energy with mpeg4-asp when NOT using Cuvid and instead software.
nevcairiel has no control over Cuvid since it's a blackbox, the DXVA2cb implementation is by himself.
aufkrawall is offline   Reply With Quote
Old 23rd July 2015, 18:07   #32087  |  Link
Akeno
Registered User
 
Join Date: Jul 2015
Location: Seattle, Washington
Posts: 53
Quote:
Originally Posted by Thunderbolt8 View Post
why, in how far is judging this different than judging it from an image which has been up- and downscaled? how can you only be sure the result you see is from the algorythm you want to look at and has not been influenced by the other one?
I'm not sure what you're trying to get at here. The nature of scaling involves throwing away details when downscaling and preserving what details still exist when upscaling. Unless you had an analog source captured at two different optical zoom levels, this is the best we can do. Most digital sources we receive are going to be upscaled or downscaled using algorithms.

Last edited by Akeno; 23rd July 2015 at 18:10.
Akeno is offline   Reply With Quote
Old 23rd July 2015, 18:35   #32088  |  Link
Ver Greeneyes
Registered User
 
Join Date: May 2012
Posts: 445
I do think the question of "which algorithm do studios/encoders most likely use to downscale footage" is an interesting question, since it will affect the look of what we're scaling and the usually unknown ground truth that we're trying to reach. I don't think it's likely to be Catmull-Rom LL AR if only because of madshi's anti-ringing algorithm - I'd imagine it would be either some form of Bicubic or Lanczos - or if they were going for a softer look maybe even Bilinear. Unfortunately that adds another variable to testing, and in the end you're just guessing.
Ver Greeneyes is offline   Reply With Quote
Old 23rd July 2015, 19:15   #32089  |  Link
Thunderbolt8
Registered User
 
Join Date: Sep 2006
Posts: 2,171
Quote:
Originally Posted by Akeno View Post
I'm not sure what you're trying to get at here. The nature of scaling involves throwing away details when downscaling and preserving what details still exist when upscaling. Unless you had an analog source captured at two different optical zoom levels, this is the best we can do. Most digital sources we receive are going to be upscaled or downscaled using algorithms.
when watching a movie on a screen of a different size, there will also only be one instance of scaling, either downscaling or upscaling. so I wonder why we dont do the comparisons that way as well. instead we use the other scaling method as well which might just influence the scaling method we want to judge in some way.
__________________
Laptop Acer Aspire V3-772g: i7-4202MQ, 8GB Ram, NVIDIA GTX 760M (+ Intel HD 4600), Windows 8.1 x64, madVR (x64), MPC-HC (x64), LAV Filter (x64), XySubfilter (x64)
Thunderbolt8 is offline   Reply With Quote
Old 23rd July 2015, 19:32   #32090  |  Link
bozokaydin
Registered User
 
Join Date: Apr 2011
Posts: 16
In my home i am watching almost 1080p high quality rip movies on 1080p 80 inch Tv. What is your suggestions for this watching type? I am not using image upcaling, downscaling and image doubling.i have core i7 cpu and gtx 690
bozokaydin is offline   Reply With Quote
Old 23rd July 2015, 20:20   #32091  |  Link
har3inger
Registered User
 
Join Date: Feb 2014
Posts: 139
Turn off every trade quality for performance tick box.

Pick something nice for chroma up scaling (nnedi3 is popular, but will make fan noise because it costs a lot)

Use one of the error diffusion dithering methods, or ordered dithering if fans are too loud.

Use smooth motion.

Set up the 3D lut for your display of you have the means to.

Play around with debanding and the source sharpener filters as you see fit.

Make sure you don't get dropped frames by checking stats with ctrl j

Enjoy.
har3inger is offline   Reply With Quote
Old 23rd July 2015, 20:36   #32092  |  Link
bozokaydin
Registered User
 
Join Date: Apr 2011
Posts: 16
Quote:
Originally Posted by har3inger View Post
Turn off every trade quality for performance tick box.

Pick something nice for chroma up scaling (nnedi3 is popular, but will make fan noise because it costs a lot)

Use one of the error diffusion dithering methods, or ordered dithering if fans are too loud.

Use smooth motion.

Set up the 3D lut for your display of you have the means to.

Play around with debanding and the source sharpener filters as you see fit.

Make sure you don't get dropped frames by checking stats with ctrl j

Enjoy.
Thank you. For Nnedi3 64 neurons chroma upsampling do i ihave to activate superres filter? Do you suggest me to use Madvr image enchantment sharpening?
__________________
HTPC:

Windows 8 64bit
Nvdia GTX 690
i7 3570k
Corsair H100 Water Cooling
Corsair 128 Gb Force SSD x 2 Raid 0+ Samsung 2 TBx 2 Hard Disc Raid 0
Creative Titanium HD
bozokaydin is offline   Reply With Quote
Old 23rd July 2015, 22:00   #32093  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,106
Quote:
Originally Posted by bozokaydin View Post
In my home i am watching almost 1080p high quality rip movies on 1080p 80 inch Tv. What is your suggestions for this watching type? I am not using image upcaling, downscaling and image doubling.i have core i7 cpu and gtx 690
These are my settings. If you have a good GPU, substituting NNEDI3 for super-xbr would make sense. A first step would be test for your display's compatibility with 10-bit output. There is a thread dedicated to this in the software forum. Based on the resolution of your display, it is most likely an 8-bit panel, but it is worth exploring, as this is a component of picture quality.

Profile: "1080p"
  • Chroma: super-xbr150 + AR
  • Image: Jinc3 + AR
  • Luma Doubling: Off
  • Upscaling Refinement: Off
  • Artifact Removal - Debanding: Medium/High
  • Image Enhancements: FineSharp (strength: 0.5)
  • Dithering: Error Diffusion 2

Use of Image Enhancements are completely up to personal taste as any additional sharpening can look artificial to many people. I have embraced sharpening mostly because I am trying to stay current with the latest madVR features. I'm not totally sold on the look, however.

I would stay away from the SuperRes chroma filter. It is not nearly as useful as SuperRes when used from Upscaling Refinement.

Last edited by Warner306; 23rd July 2015 at 23:13.
Warner306 is offline   Reply With Quote
Old 23rd July 2015, 23:12   #32094  |  Link
Warner306
Registered User
 
Join Date: Dec 2014
Posts: 1,106
Now that Kodi DSPlayer has integrated low latency mode, I have noticed a new issue. I believe this addition has made 1080p60 playback a little stuttery with some noticeable blurring on moving objects. It is subtle, but I cannot detect the same problem at 1080p24.

I have switched between MPC-BE and DSPlayer at 60hz and find DSPlayer is now less smooth - the blurring on objects being particularly noticeable.

Was 3/2 pulldown properly tested when low latency mode was added? Something does not look right at 1080p60 when played through DSPlayer.

Last edited by Warner306; 23rd July 2015 at 23:54.
Warner306 is offline   Reply With Quote
Old 24th July 2015, 00:11   #32095  |  Link
bozokaydin
Registered User
 
Join Date: Apr 2011
Posts: 16
Quote:
Originally Posted by Warner306 View Post
These are my settings. If you have a good GPU, substituting NNEDI3 for super-xbr would make sense. A first step would be test for your display's compatibility with 10-bit output. There is a thread dedicated to this in the software forum. Based on the resolution of your display, it is most likely an 8-bit panel, but it is worth exploring, as this is a component of picture quality.

Profile: "1080p"
  • Chroma: super-xbr150 + AR
  • Image: Jinc3 + AR
  • Luma Doubling: Off
  • Upscaling Refinement: Off
  • Artifact Removal - Debanding: Medium/High
  • Image Enhancements: FineSharp (strength: 0.5)
  • Dithering: Error Diffusion 2

Use of Image Enhancements are completely up to personal taste as any additional sharpening can look artificial to many people. I have embraced sharpening mostly because I am trying to stay current with the latest madVR features. I'm not totally sold on the look, however.

I would stay away from the SuperRes chroma filter. It is not nearly as useful as SuperRes when used from Upscaling Refinement.
Yes my display is 8 bit. I am very pleased with the settings you offer for 1080p Bluray content. Thank you.
__________________
HTPC:

Windows 8 64bit
Nvdia GTX 690
i7 3570k
Corsair H100 Water Cooling
Corsair 128 Gb Force SSD x 2 Raid 0+ Samsung 2 TBx 2 Hard Disc Raid 0
Creative Titanium HD
bozokaydin is offline   Reply With Quote
Old 24th July 2015, 00:19   #32096  |  Link
har3inger
Registered User
 
Join Date: Feb 2014
Posts: 139
Quote:
Originally Posted by bozokaydin View Post
Thank you. For Nnedi3 64 neurons chroma upsampling do i ihave to activate superres filter? Do you suggest me to use Madvr image enchantment sharpening?
You probably shouldn't activate the superres filter until it is developed a little more. In its current form it's not so good. nnedi3 64 neurons is also quite overkill for chroma upscaling, but since you have the GTX 690 to run it, may as well if the fan noise is tolerable.

SuperXBR is being actively developed like Warner mentions, and is a very cheap alternative to nnedi3 that still looks almost as good. Many people here would probably say that it's 2nd place for the best quality upscaler. You can play with the sharpness of that algorithm a bit as well. I find 150 can cause reds to pop unnaturally in a scene and use 100 instead, but it's basically whatever you think looks best. Sharper is noticeably better for animated content.

You can try out all the image enhancements for yourself easily by pausing a video, then turning on/off the settings to see what they do. It's all personal preference, so you need to see for yourself and come up with a combo that you find pleasing.

As a very general overview:

Finesharp enhances details and artifacts in the source. It can look good with exceptionally clean sources, but has a tendency to bring out ringing, blocking, and other artifacts (which are inevitable with almost all formats of compressed video with sane bitrates).

Lumasharpen is similar to what some games use for edge sharpening. It's a different look compared to finesharp. You might like it, you might not.

Adaptive sharpen tries to sharpen details more than artifacts in the source. In practice, it introduces some artifacts of its own sometimes. Again, try it yourself to see if it improves the image in your opinion.

Personally I use none of these and prefer a "purist" approach where I try to reproduce an image accurately instead of favoring heavy processing like sharpen filters.
har3inger is offline   Reply With Quote
Old 24th July 2015, 00:39   #32097  |  Link
Akeno
Registered User
 
Join Date: Jul 2015
Location: Seattle, Washington
Posts: 53
Quote:
Originally Posted by Thunderbolt8 View Post
when watching a movie on a screen of a different size, there will also only be one instance of scaling, either downscaling or upscaling. so I wonder why we dont do the comparisons that way as well. instead we use the other scaling method as well which might just influence the scaling method we want to judge in some way.
Again, I'm not sure what you're getting at. We are trying to achieve scaling that is closest to the image we have designated as groundtruth. The grountruth is what the image should look like after upscaling. We need to scale down the groundtruth in this case to produce an image that we will scale back up to approach the groundtruth unless you have a source that is identical except for size differences. If I'm wrong in assuming what you're saying, please feel free to help clarify.
Akeno is offline   Reply With Quote
Old 24th July 2015, 03:51   #32098  |  Link
avinab
Registered User
 
Join Date: May 2015
Posts: 14
Not able to change to NNEDI3

In build 88.20 when i am selecting NNEDI3 in chroma upscaling,it is showing JINC in OSD.Previously If i select NNEDI3 in chroma or image doubling MADVR crashes,but now its not crashing but shows JINC in OSD instead of NNEDI3.But it shows super-xbr,NEDI when i select those.I am on windows 10 build 10240.Nvidia driver 353.53 Geforce GTX 980.
avinab is offline   Reply With Quote
Old 24th July 2015, 08:09   #32099  |  Link
michkrol
Registered User
 
Join Date: Nov 2012
Posts: 167
I'll assume NNEDI3 has never worked for you on (this installation of) Windows 10.
You are probably missing OpenCL, so I'd recommend installing drivers directly from nvidia's website.
The newest I could find is a hotfix: https://forums.geforce.com/default/t...eased-7-8-15-/
michkrol is offline   Reply With Quote
Old 24th July 2015, 09:44   #32100  |  Link
Gravitator
Registered User
 
Join Date: May 2014
Posts: 169
All kind people hello!
Version 0.88.16 during pause 0% GPU (MPC-HC 1.7.9.54);
Version 0.88.20 when paused consumes 40% of the GPU. - Перемудрил Мад;
(Win x64, Pentium E6600, nVidia GTX750 (DR 350.12).
Gravitator 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 09:15.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.