[FFmpeg-devel] How to fill the pts field of AVFRAME struct when encoding a frame

Tingustar tingustar
Sat Mar 22 16:59:25 CET 2008

Hi all,

I'm using FFMPEG's mpeg4 codec to encode live video and then stream it by network. Here the question is that when calling avcodec_encode_video(), does I need to fill the pts field of AVFRAME struct or just keep it AV_NOPTS_VALUE? There is only video frame in my application so audio-video sync is not needed. In this case,I think pts is not useful (even not necessary), is it right? I'm thinking in the receive side, the player(written by myself) can correctly re-construct the stream only by a per-VOP sequence number instead of pts(timestamp).Thank you very much for your answers.



More information about the ffmpeg-devel mailing list