I'm debugging a problem importing mp4 into fcpx (final cut pro X). I
backtracked it to m2ts re-containerizing. I don't think this question about
mkv gets to the problem, but it is the first discrepancy in the entire
Is there any chance that ffmpeg adds extra blank frames to the end of a
video stream when changing container from m2ts->mkv?
('ffmpeg -i foo.m2ts -c copy -map 0 foo.mkv')
For a specific video, if I do this passthrough conversion with ffmpeg, the
video in the mkv has 189280 frames.
If I do this passthrough conversion with dvdfab (a commercial tool), the
video in the mkv has 189260 frames.
I'm doing both conversions on Mac, but with avisynth on Windows, I have
confirmed that the videos are frame-by-frame identical except for the extra
20 blank frames at the end of the ffmpeg version.
The extra 20 frames don't force a whole second duration alignment.
The audio runs 189280.x frames, ie less than one frame longer than the
extra padding. Is ffmpeg padding out extra video frames to be within a
frame of the audio length? If so, why does it do that? Note that ffmpeg
adds this padding even if I don't map the audio tracks to the mkv.
ffmpeg-user mailing list
[hidden email] https://ffmpeg.org/mailman/listinfo/ffmpeg-user
To unsubscribe, visit link above, or email
[hidden email] with subject "unsubscribe".