On 04/11/14 15:38, Martin Storsjö wrote:
This should avoid potential small negative offsets due to rounding error. --- libavformat/mux.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)diff --git a/libavformat/mux.c b/libavformat/mux.c index 4067c16..e4ba209 100644 --- a/libavformat/mux.c +++ b/libavformat/mux.c @@ -327,7 +327,8 @@ static int write_packet(AVFormatContext *s, AVPacket *pkt) s->offset_timebase = time_base; } if (s->offset) - offset = av_rescale_q(s->offset, s->offset_timebase, time_base); + offset = av_rescale_q_rnd(s->offset, s->offset_timebase, time_base, + AV_ROUND_UP); if (pkt->dts != AV_NOPTS_VALUE) pkt->dts += offset;
Ok, I guess. _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
