skampy
4th March 2011, 14:29
I've had problems in the past with broken demuxed video streams inside mkv containers in which some would stop completely and go back to the start of the video when it reached a certain point, and others where the video would simply 'jump' a small amount of time ahead which was nearly always proceeded by corrupt frames. Since I've had just that latter problem with a recent encode of mine, I thought I'd ask the experts what could cause this problem?
To give you a visual idea of what I mean by 'corrupt' frames, have a look at this screen:
http://i1208.photobucket.com/albums/cc362/follz20/Other/f90437.png
This frame is part of a corrupt video sequence of about 10 seconds which was proceeded by a small, 30 second 'jump' of the video.
My method of encoding, which is successful probably 98% of the time, is to demux the video & audio streams with eac3to and mux them back into a mkv container with mmg (4.4.0. with NO header stripping and which never gives me any errors even with corrupt mkv's) before I create my AVS script. The reason I don't create the script directly from the m2ts is because in the past, I've had horrendous problems with corrupt frames (lol), video gaps/jumps and other video artefacts with the finished result.
(On that note, if anyone has a combination of filters that work for them when encoding directly from the m2ts blu-ray files, post a reply or PM me. Cheers! :) )
So, my question basic boils down to what causes this and why? Does the corruption come from eac3to when the raw video file is demuxed or from mmg when it is remuxed (x264 obviously isn't the problem here)? And, more importantly, is there any way to check the final remux of an encode for gaps/jumps in the video stream (is the -check function in eac3to useful for this?)?
Cheers for any help/info :thanks:
To give you a visual idea of what I mean by 'corrupt' frames, have a look at this screen:
http://i1208.photobucket.com/albums/cc362/follz20/Other/f90437.png
This frame is part of a corrupt video sequence of about 10 seconds which was proceeded by a small, 30 second 'jump' of the video.
My method of encoding, which is successful probably 98% of the time, is to demux the video & audio streams with eac3to and mux them back into a mkv container with mmg (4.4.0. with NO header stripping and which never gives me any errors even with corrupt mkv's) before I create my AVS script. The reason I don't create the script directly from the m2ts is because in the past, I've had horrendous problems with corrupt frames (lol), video gaps/jumps and other video artefacts with the finished result.
(On that note, if anyone has a combination of filters that work for them when encoding directly from the m2ts blu-ray files, post a reply or PM me. Cheers! :) )
So, my question basic boils down to what causes this and why? Does the corruption come from eac3to when the raw video file is demuxed or from mmg when it is remuxed (x264 obviously isn't the problem here)? And, more importantly, is there any way to check the final remux of an encode for gaps/jumps in the video stream (is the -check function in eac3to useful for this?)?
Cheers for any help/info :thanks: