On Wed, Oct 12, 2016 at 8:04 AM, Carl Eugen Hoyos <ceffm...@gmail.com> wrote:
> 2016-10-12 16:03 GMT+02:00 Thomas Worth <d...@rarevision.com>: > > On Wed, Oct 12, 2016 at 2:50 AM, Carl Eugen Hoyos <ceffm...@gmail.com> > > wrote: > > > >> 2016-10-12 9:31 GMT+02:00 Thomas Worth <d...@rarevision.com>: > >> > >> > Has anyone actually looked at the MP4 file, BrokenVideo-8min.mp4? > > > Just look at mdhd and you'll see what I'm talking about. > > This is the second comment today that could be misinterpreted: > Please remember that not everybody is a native speaker and > be more careful. > Understood. Thanks. Allow me a question: > Did you test with any non-QT based player? > If your analysis were right, nothing could play the sample, so I > believe it is safe to say that your analysis cannot be correct. > Yes. Tested the file with Adobe After Effects version 13 on a Mac, and it failed. I am pretty sure After Effects uses its own media decoder for MOV/MP4 and not QuickTime. The mdhd atom in the file is version 1 and has 64bit time and > duration fields and is therefore twelve bytes larger than a > version 0 mdhd atom. > Right. I took another look at this, and QuickTime Player on OS X 10.12 does report a 23.976 frame rate. I re-encoded Alexey's file with my copy of ffmpeg using libx264 and using -video_track_timescale 10000000, and it wrote an MP4 file with the same mdhd atom as Alexey's file. Obviously, the duration field of the mdhd atom (version 1) needed to be 64 bit since the video was so long that the PTS values would have exceeded the 32 bit unsigned limit. In this case, since the file has 11509 frames, the last PTS would be 11509*417083, or 4800208247. Anyway, it worked with QuickTime Player so we can rule this out. It may not be a container problem after all as I originally suspected, but an H264 decoder problem. If the file were encoded as JPEG and not H.264 and it works, then the problem is codec related. Thomas _______________________________________________ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".