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 11th November 2013, 13:55   #20841  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
This just confuses me even more. So running both with TV output levels the image looks the same with 0.85.3 and 0.85.4? But running both with PC output levels, you get different brightness? This doesn't make any sense to me at all. Can you please make screenshots with 0.85.3, 0.85.4 and 0.86.4, each with both TV and PC levels? So 6 screenshots overall? To be honest, I'm near to giving up on this, because it doesn't make ANY sense to me whatsoever, from a technical point of view. Especially since you say it only occurs with one specific video, which makes it even more weird.
madshi is offline   Reply With Quote
Old 11th November 2013, 14:48   #20842  |  Link
andybkma
Registered User
 
Join Date: Sep 2006
Posts: 212
Win7 SP1, Zoom Player Max, mVR Deband 14, OPtimus Graphics HD4000 with Nvidia 650M

Been using mVR Deband 14 since yesterday and I have noticed now that about 50% of the time when coming out of FSE back into windowed mode I get the following error code displayed in the top left corner of ZP:

"madVR reports: resetting Direct3D device failed 8876017c"

I press stop on ZP and the error resets itself (error dialogue disappears) and ZP (and mVR) can be used again. How can I get you an error report? It doesn't duplicate all the time...
andybkma is offline   Reply With Quote
Old 11th November 2013, 15:34   #20843  |  Link
Werewolfy
Registered User
 
Join Date: Feb 2013
Posts: 137
@madshi
I did some tests with your latested build. The low preset is fine. nonGradientPenalty removal concerned me a little at first but since you decreased Gradient (or MidDif now), it seems we have the same result here. Anyway, after you created AngleBoost I didn't really feel that nonGradientPenalty was very useful.

I have one little complain about the medium preset. MaxAngle at 0.22 is a little too high for me. Did you find a video where MaxAngle at 0.18 or even 0.20 was too low? It's true that I had set MaxAngle at 0.25 for the high preset but it was only useful for bad videos. I don't think the medium preset needs this because we lose some details sometimes and this preset is too weak to make a difference for bad videos, MaxAngle at 0.22 doesn't seem to change anything for debanding here.

The high preset seems fine to me. I was a little surprised you desactivated AngelBoost here but after making some tests, it makes sense.

I think "Artifact removal" is not a really good name for this function. I understand why you don't call it "debanding" but I don't know... "Postprocessing", "Filter" or even "Postprocessing filter" or "Smoothing" doesn't fit more than "Artifact removal"?

I also tested your new feature called "strength during fade in/out" . If you set it on high, when you have a fade in/out your lose a lot of details during this short period of time. The fade in/out seems to last a little bit longer because details make more time to appear. IMHO, medium seems a better choice by default.
Werewolfy is offline   Reply With Quote
Old 11th November 2013, 15:59   #20844  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by andybkma View Post
Been using mVR Deband 14 since yesterday and I have noticed now that about 50% of the time when coming out of FSE back into windowed mode I get the following error code displayed in the top left corner of ZP:

"madVR reports: resetting Direct3D device failed 8876017c"
And this does not happen with the official release build?

Quote:
Originally Posted by Werewolfy View Post
I have one little complain about the medium preset. MaxAngle at 0.22 is a little too high for me. Did you find a video where MaxAngle at 0.18 or even 0.20 was too low? It's true that I had set MaxAngle at 0.25 for the high preset but it was only useful for bad videos. I don't think the medium preset needs this because we lose some details sometimes and this preset is too weak to make a difference for bad videos, MaxAngle at 0.22 doesn't seem to change anything for debanding here.
I've no problem at all lowering the MaxAngle for the medium preset. Which value would you prefer?

Quote:
Originally Posted by Werewolfy View Post
I think "Artifact removal" is not a really good name for this function. I understand why you don't call it "debanding" but I don't know... "Postprocessing", "Filter" or even "Postprocessing filter" or "Smoothing" doesn't fit more than "Artifact removal"?
I don't want to use one separate settings page for every algorithm. Otherwise I would have chosen "debanding" as the page name. Some day I might add other algorithms, e.g. deblocking or deringing, or things like that. Or even noise reduction, as suggested by some. I don't know if and when I might add those algorithms. But if I do, all of them fit nicely under the name "artifact removal", because all of those algorithms try to fix/repair/improve a bad source or a bad encoding. On the other hand, sharpening is something completely different. And if I ever add sharpening or detail enhancement processing, that would get its own page in the settings dialog. All the algorithms on the "artifact removal" page will be performed *before* scaling. Sharpening and detail enhancement will probably be performed *after* scaling. It's simply something completely different. If you have a better suggestion for the artifact removal settings page name, let me know. But "post processing" doesn't fit, because that would apply to sharpening, too. And "smoothing" is specific to debanding, so it's a bad name, too.

Quote:
Originally Posted by Werewolfy View Post
I also tested your new feature called "strength during fade in/out" . If you set it on high, when you have a fade in/out your lose a lot of details during this short period of time. The fade in/out seems to last a little bit longer because details make more time to appear. IMHO, medium seems a better choice by default.
The fade in/out detection should not last a single frame too long. I'm pretty sure it works correctly. Please double check.

I would be ok with choosing medium as default, if the majority of "expert users" here agree on that. Personally, I think "high" is the better choice, because IMHO during fade ins/outs lack of banding artifacts is more important than losing a tiny bit of detail, but that's just my personal opinion, of course. I'll bow to the majority vote.
madshi is offline   Reply With Quote
Old 11th November 2013, 16:28   #20845  |  Link
andybkma
Registered User
 
Join Date: Sep 2006
Posts: 212
Quote:
Originally Posted by madshi View Post
And this does not happen with the official release build?
Nope, never got that error till yesterday when started using the deband version, 13 then 14. Hopefully I can get the problem to duplicate more frequently...
andybkma is offline   Reply With Quote
Old 11th November 2013, 16:32   #20846  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Does reducing the number of windowed mode backbuffers and/or the number of exclusive mode prepresented frames help in any way?
madshi is offline   Reply With Quote
Old 11th November 2013, 17:35   #20847  |  Link
Werewolfy
Registered User
 
Join Date: Feb 2013
Posts: 137
Quote:
Originally Posted by madshi View Post
I've no problem at all lowering the MaxAngle for the medium preset. Which value would you prefer?
I would set it at 0.18.

Quote:
Originally Posted by madshi View Post

I don't want to use one separate settings page for every algorithm. Otherwise I would have chosen "debanding" as the page name. Some day I might add other algorithms, e.g. deblocking or deringing, or things like that. Or even noise reduction, as suggested by some. I don't know if and when I might add those algorithms. But if I do, all of them fit nicely under the name "artifact removal", because all of those algorithms try to fix/repair/improve a bad source or a bad encoding. On the other hand, sharpening is something completely different. And if I ever add sharpening or detail enhancement processing, that would get its own page in the settings dialog. All the algorithms on the "artifact removal" page will be performed *before* scaling. Sharpening and detail enhancement will probably be performed *after* scaling. It's simply something completely different. If you have a better suggestion for the artifact removal settings page name, let me know. But "post processing" doesn't fit, because that would apply to sharpening, too. And "smoothing" is specific to debanding, so it's a bad name, too.
Ok I understand why you've chosen this name. I don't have a suggestion right now but I will think about it. Anyway it's not very important.

Quote:
Originally Posted by madshi View Post
The fade in/out detection should not last a single frame too long. I'm pretty sure it works correctly. Please double check.

I would be ok with choosing medium as default, if the majority of "expert users" here agree on that. Personally, I think "high" is the better choice, because IMHO during fade ins/outs lack of banding artifacts is more important than losing a tiny bit of detail, but that's just my personal opinion, of course. I'll bow to the majority vote.
On this sample posted earlier, I find the face takes longer to appear when high is selected. Maybe it's a too extreme example? https://dl.dropboxusercontent.com/u/...nshin_-_01.mkv
Werewolfy is offline   Reply With Quote
Old 11th November 2013, 17:53   #20848  |  Link
Demonik
Registered User
 
Join Date: May 2012
Posts: 10
Quote:
Originally Posted by madshi View Post
This just confuses me even more. So running both with TV output levels the image looks the same with 0.85.3 and 0.85.4? But running both with PC output levels, you get different brightness? This doesn't make any sense to me at all. Can you please make screenshots with 0.85.3, 0.85.4 and 0.86.4, each with both TV and PC levels? So 6 screenshots overall? To be honest, I'm near to giving up on this, because it doesn't make ANY sense to me whatsoever, from a technical point of view. Especially since you say it only occurs with one specific video, which makes it even more weird.
That is correct. screens here:

0.85.3PC 0.85.4PC 0.86.4PC
0.85.3TV 0.85.4TV 0.86.4TV
Demonik is offline   Reply With Quote
Old 11th November 2013, 18:16   #20849  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
And 0.86.3 behaves like 0.85.4? The only way that makes sense is if the "source black level" and/or "source white level" controls would be set to some unexpected values. These controls can be modified by defining custom key shortcuts in the madVR settings dialog. Or by using file name tagging. I can't imagine you accidently modified the controls, and the file name "sample002.mkv" doesn't contain any file name tags. So it all still makes no sense to me. Try assigning a keyboard shortcut "source color controls - reset" and press that. Does that change anything?

What is the full file name and file path of this video file on your PC?

Are you using any custom pixel shaders?
madshi is offline   Reply With Quote
Old 11th November 2013, 18:31   #20850  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by Werewolfy View Post
I would set it at 0.18.
Ok.

Quote:
Originally Posted by Werewolfy View Post
On this sample posted earlier, I find the face takes longer to appear when high is selected. Maybe it's a too extreme example? https://dl.dropboxusercontent.com/u/...nshin_-_01.mkv
It's a pretty mean example, but I can see what you mean. When frame stepping through the fade in, the first few frames are quite washed out with the "high" setting. The remaining frames of the fade in show little difference between "low" and "high", though. And in motion I'm not sure I see much of a difference, either. There's probably always a sample which works bad with a different setting. I could probably find a sample which with "medium" setting would still show banding during a fade in/out while it wouldn't show banding with "high". With such a sample "high" would look better. So which is the better overall setting? I'm not really sure.

Opinions from other users welcome, as well.
madshi is offline   Reply With Quote
Old 11th November 2013, 18:41   #20851  |  Link
Demonik
Registered User
 
Join Date: May 2012
Posts: 10
Yes, 0.86.3 behaves like 0.85.4, i've assigned a key to source color control reset and .. it works, i get a normal picture on the "bad" versions. I haven't made any other settings, all were default. weird.
Ok. tested the filename and path thingie. the files were in a folder named "Men in Black 1997" turns out that if i change the 1997 to some other value it changes the source .. black level?

Last edited by Demonik; 11th November 2013 at 18:56.
Demonik is offline   Reply With Quote
Old 11th November 2013, 19:16   #20852  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
OMG, yes, "Black 1997" gets treated as a valid tag, changing the black level of the video to 1997. Obviously that doesn't make much sense. But any "someTag someValue" combination is accepted as a file name tag. And "black" is a supported tag name. If you put the 1997 in brackets like "Men in Black (1997)", the problem should go away. That's also how IMDB names movies and how I'm storing movies on my server. Well, I'm glad we got this mystery solved.

Now the question is: Should I require a "tag=value" syntax? Currently I support "tag=value", "tag:value" and "tag value". And "black" is a supported tag.
madshi is offline   Reply With Quote
Old 11th November 2013, 19:30   #20853  |  Link
Demonik
Registered User
 
Join Date: May 2012
Posts: 10
Well that is not up to me but i'm also glad we solved this one.
It's not hard for me (now) to reset source color controls if i'm forced to play a file from such a folder.
Thank you so much!
Demonik is offline   Reply With Quote
Old 11th November 2013, 20:03   #20854  |  Link
Qotscha
Registered User
 
Join Date: Dec 2012
Posts: 40
Quote:
Originally Posted by madshi View Post
Now the question is: Should I require a "tag=value" syntax? Currently I support "tag=value", "tag:value" and "tag value". And "black" is a supported tag.
I think it would be good to require "tag=value" syntax to avoid confusions. For example, I use file names like "Program name 2013.11.11 21.00" for TV recordings, so requiring "tag=value" syntax would decrease the probability of encountering problems with accidental tags quite much.
Qotscha is offline   Reply With Quote
Old 11th November 2013, 21:27   #20855  |  Link
DragonQ
Registered User
 
Join Date: Mar 2007
Posts: 934
Quote:
Originally Posted by madshi View Post
OMG, yes, "Black 1997" gets treated as a valid tag, changing the black level of the video to 1997. Obviously that doesn't make much sense. But any "someTag someValue" combination is accepted as a file name tag. And "black" is a supported tag name. If you put the 1997 in brackets like "Men in Black (1997)", the problem should go away. That's also how IMDB names movies and how I'm storing movies on my server. Well, I'm glad we got this mystery solved.

Now the question is: Should I require a "tag=value" syntax? Currently I support "tag=value", "tag:value" and "tag value". And "black" is a supported tag.
Hahaha.

Would it not be more sensible to only scan for tags inside square brackets? For example:

Men In Black [1997, 1080p24, 5.1]
__________________
TV Setup: LG OLED55B7V; Onkyo TX-NR515; ODroid N2+; CoreElec 9.2.7
DragonQ is offline   Reply With Quote
Old 11th November 2013, 21:52   #20856  |  Link
XeoneR
Registered User
 
Join Date: Jul 2003
Posts: 17
Quote:
Originally Posted by madshi View Post
Have you tried the different rendering modes (windowed, overlay, fullscreen exclusive mode)?
Is the CPU consumption equally high in all of them?
Overlay isn't available to me. But in general yes.

It's getting better though.
I've overclocked my stuff and checked drivers.
Also i've completely solved the previous issues.

Thank you for the best renderer and your kind support.
XeoneR is offline   Reply With Quote
Old 11th November 2013, 22:48   #20857  |  Link
turbojet
Registered User
 
Join Date: May 2008
Posts: 1,840
Found a video that crashes with deint=ivtc and potplayer but 'send bug report' opens a blank browser window with mailto: but nothing happens. Is it reporting?

Concerning name tags, : isn't allowed in windows filenames afaik. = is such an uncommon tag that it makes sense. [] are fairly common for year.

Would it make sense to deband after downsizing to lower gpu load?
__________________
PC: FX-8320 GTS250 HTPC: G1610 GTX650
PotPlayer/MPC-BE LAVFilters MadVR-Bicubic75AR/Lanczos4AR/Lanczos4AR LumaSharpen -Strength0.9-Pattern3-Clamp0.1-OffsetBias2.0
turbojet is offline   Reply With Quote
Old 11th November 2013, 23:04   #20858  |  Link
Werewolfy
Registered User
 
Join Date: Feb 2013
Posts: 137
Quote:
Originally Posted by madshi View Post
It's a pretty mean example, but I can see what you mean. When frame stepping through the fade in, the first few frames are quite washed out with the "high" setting. The remaining frames of the fade in show little difference between "low" and "high", though. And in motion I'm not sure I see much of a difference, either. There's probably always a sample which works bad with a different setting. I could probably find a sample which with "medium" setting would still show banding during a fade in/out while it wouldn't show banding with "high". With such a sample "high" would look better. So which is the better overall setting? I'm not really sure.

Opinions from other users welcome, as well.
In motion I can see the difference between medium and high. The face appears without any details and is very smooth with high.

But I have a very weird issue with this new feature. When I'm watching a movie in 24hz, I get frames dropped when I activate a different strength for fade in/out. It happens only in exclusive mode. It doesn't happen in windowed mode or in 60hz mode.
Werewolfy is offline   Reply With Quote
Old 12th November 2013, 00:23   #20859  |  Link
Boltron
Registered User
 
Boltron's Avatar
 
Join Date: May 2011
Posts: 94
Quote:
Originally Posted by DragonQ View Post
Hahaha.

Would it not be more sensible to only scan for tags inside square brackets? For example:

Men In Black [1997, 1080p24, 5.1]
Noooo, I use [] in my naming convention.
Boltron is offline   Reply With Quote
Old 12th November 2013, 01:18   #20860  |  Link
MistahBonzai
Registered User
 
Join Date: Mar 2013
Posts: 101
Frame drop reporting with 'fade' limmited to FSE.

Quote:
Originally Posted by madshi View Post
I can reproduce the problem. However, playback still seems perfectly smooth on my PC. It's possible that this is a false alarm in the debug OSD. I'll investigate.
Yup..I don't have visual indication of actual dropped frames. CPU useage dosen't increase and it's limmited to FSE mode.
MistahBonzai 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 06:23.


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