decoder/ffmpeg: don't require key frame for seeking
Use flag AV_TIME_BASE.
This commit is contained in:
parent
42d8c2981f
commit
724a59aaf7
1
NEWS
1
NEWS
@ -4,6 +4,7 @@ ver 0.16.5 (2010/??/??)
|
||||
- rewind: reduce heap usage
|
||||
* decoder:
|
||||
- ffmpeg: higher precision timestamps
|
||||
- ffmpeg: don't require key frame for seeking
|
||||
|
||||
|
||||
ver 0.16.4 (2011/09/01)
|
||||
|
@ -447,7 +447,8 @@ ffmpeg_decode(struct decoder *decoder, struct input_stream *input)
|
||||
int64_t where =
|
||||
decoder_seek_where(decoder) * AV_TIME_BASE;
|
||||
|
||||
if (av_seek_frame(format_context, -1, where, 0) < 0)
|
||||
if (av_seek_frame(format_context, -1, where,
|
||||
AV_TIME_BASE) < 0)
|
||||
decoder_seek_error(decoder);
|
||||
else
|
||||
decoder_command_finished(decoder);
|
||||
|
Loading…
Reference in New Issue
Block a user