diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2014-01-21 03:30:25 +0100 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2014-01-21 04:10:44 +0100 |
| commit | e9b836cd87c2404f4d2d7492ddf982dd446076f8 (patch) | |
| tree | 3fb104a322646209a2672dbc2bed088e55f29c36 /tests | |
| parent | 13e0109a5f2c452f29da970a9acab06c3ce07833 (diff) | |
| parent | d01e684186bc1631bc176f06b89d33c27ec0d24d (diff) | |
| download | ffmpeg-e9b836cd87c2404f4d2d7492ddf982dd446076f8.tar.gz | |
Merge commit 'd01e684186bc1631bc176f06b89d33c27ec0d24d'
* commit 'd01e684186bc1631bc176f06b89d33c27ec0d24d':
mov: do not set avg_frame_rate in the demuxer
Conflicts:
libavformat/mov.c
Not merged
The default fallback code that sets avg_frame_rate guesses the value
from a very small number of frames at the videos start and is wrong in
some cases for the file as a whole.
Now that makes it guess vs guess, but the code in mov is much faster
Also ideally the mov code should be fixed and give a exact avg_frame_rate
based on the actual cts from the index (CTTS/STTS)
If someone has sample(s) which show wrong fps values, please open a ticket
on trac or mail me if its private.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'tests')
0 files changed, 0 insertions, 0 deletions