Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 05-04-10, 11:10 AM   #13
cehoyos
FFmpeg developer
 
Join Date: Jan 2009
Location: Vienna, Austria
Posts: 467
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Originally Posted by Stu-e View Post
Am I correct assuming progressive and interlace flagging reliability is generally quite poor?
I tried to explain I suspect it is unreliable (for our purpose: I am sure the provider will explain why he thinks the flags are set correctly, be it because he doesn't use field-encoding or because the original progressive 25fps material was 2:2 telecined to provide "interlaced" 50 fps material) - if you can convince me I'm wrong, I will try to implement support in MPlayer.
Quote:
Can you recommend any software tools freely available under linux for analysing dvb streams to check for the presence of such flags?
I believe I used dvbsnoop once for that purpose.
Quote:
Are things more reliable for h264 streams? Reading around I get the impression there are no such flags. Instead field coding is implied from the h264 encoding method?
But since progressive material is (or at least may be) encoded in fields for 1080i we are back at the start...

Carl Eugen
cehoyos is offline   Reply With Quote
Old 05-06-10, 05:54 AM   #14
Stu-e
Registered User
 
Join Date: Apr 2010
Posts: 9
Default Re: VDPAU 2:2 pulldown detection failure

Cehoyos
I suspect my LCD TV at home analyses video to detect 2:2 pulldown and disables deinterlacing as appropriate. Am I correct in assuming with Mplayer the 'pullup' filter is similar?

Could I use this command to play video and apply yadif deinterlacing only where necessary:

mplayer -vf pullup,yadif=1 tvbroadcastfile.ts

?

Stu-e
Stu-e is offline   Reply With Quote
Old 05-06-10, 06:42 AM   #15
cehoyos
FFmpeg developer
 
Join Date: Jan 2009
Location: Vienna, Austria
Posts: 467
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Originally Posted by Stu-e View Post
Could I use this command to play video and apply yadif deinterlacing only where necessary:

mplayer -vf pullup,yadif=1 tvbroadcastfile.ts
It is possible that you know more than I do, but I don't see how pullup is supposed to disable yadif.

Carl Eugen
cehoyos is offline   Reply With Quote
Old 05-06-10, 07:17 AM   #16
Stu-e
Registered User
 
Join Date: Apr 2010
Posts: 9
Default Re: VDPAU 2:2 pulldown detection failure

Ok That was a wild guess on my part.

To summarise:
Disabling deinterlace for 2:2 pulldown video is a difficult thing to do it seems. Stream flags are unreliable in some peoples experience. VDPAU IVTC does not perform well enough in my opinion (should disable deinterlace where there is little to no movement but unfortunataly will fallback to enabled). Mplayer has no link between the pullup filter and deinterlacers like yadif.

Stu-e
Stu-e is offline   Reply With Quote
Old 05-06-10, 08:32 AM   #17
cehoyos
FFmpeg developer
 
Join Date: Jan 2009
Location: Vienna, Austria
Posts: 467
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Originally Posted by Stu-e View Post
Stream flags are unreliable
Did you test this?

Carl Eugen
cehoyos is offline   Reply With Quote
Old 05-06-10, 09:20 AM   #18
crisalide
Registered User
 
Join Date: Dec 2008
Posts: 173
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Stream flags are unreliable in some peoples experience.
At least, it's true for a lot of mpeg2 streams.
Not sure about h264.
xine-lib-1.2 vdpau support has an option for this: vdpau_honor_progressive.
Feel free to give it a try, but do not expect any magic since then we rely on provider's encoder and my guess is that the stream will be flagged as interlace
crisalide is offline   Reply With Quote
Old 01-20-11, 12:05 PM   #19
nmi
Registered User
 
Join Date: Jun 2006
Posts: 2
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Originally Posted by Stephen Warren View Post
VDPAU's inverse telecine algorithm should work fine with 2:2 in general. The issue is that in this particular clip, there is little or no movement, and hence there's little to no difference between fields from different frames/pictures.
How about sharp videos with movement? Could you get the inverse telecine algorithm to detect these cases for example:
http://ubuntuone.com/p/ZFA/parkrun_720p25.mkv
http://ubuntuone.com/p/ZF9/stockholm_720p25.mkv

Pullup does work if I smooth the vertical resolution slightly by downscaling and upscaling:
Code:
mplayer -vf scale=1280:700,scale=1280:720 -vo vdpau:deint=4:pullup stockholm_720p25.mkv
or a bit more for parkrun:
Code:
mplayer -vf scale=1280:600,scale=1280:720 -vo vdpau:deint=4:pullup parkrun_720p25.mkv
nmi is offline   Reply With Quote
Old 01-21-11, 03:53 AM   #20
cehoyos
FFmpeg developer
 
Join Date: Jan 2009
Location: Vienna, Austria
Posts: 467
Default Re: VDPAU 2:2 pulldown detection failure

Very useful test-cases!
Quote:
Pullup does work if I smooth the vertical resolution slightly by downscaling and upscaling
Are you sure?
I mean, do you see a difference between scale & deint & pullup and scale & deint & nopullup? I believe the scaling just removes the de-interlacing artefacts.

Could somebody test this with (very) old driver versions? I did test pullup once successfully...

Carl Eugen

Last edited by cehoyos; 01-21-11 at 05:11 AM. Reason: Try to clarify
cehoyos is offline   Reply With Quote

Old 01-21-11, 12:05 PM   #21
nmi
Registered User
 
Join Date: Jun 2006
Posts: 2
Default Re: VDPAU 2:2 pulldown detection failure

Quote:
Originally Posted by cehoyos View Post
I mean, do you see a difference between scale & deint & pullup and scale & deint & nopullup? I believe the scaling just removes the de-interlacing artefacts.
No, pullup definitely works after the smoothing trick and it's clearly visible in both samples. The videos are still pretty sharp after minor bicubic down- and upscaling.

Quote:
Could somebody test this with (very) old driver versions? I did test pullup once successfully...
It works most of the time, but sharp sequences seem to be problematic because the algorithm misidentifies them as interlaced. I see the same issue in some detailed Blu-ray sources.
nmi is offline   Reply With Quote
Old 01-24-11, 12:38 PM   #22
Stephen Warren
Moderator
 
Stephen Warren's Avatar
 
Join Date: Aug 2005
Posts: 1,327
Default Re: VDPAU 2:2 pulldown detection failure

nmi,

Thanks for the samples. I've filed a bug to investigate them, but haven't yet had a chance to confirm the problem locally.
Stephen Warren is offline   Reply With Quote
Reply


Thread Tools

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 -5. The time now is 03:57 AM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.