While I'm using av_read_frame to extract frames from some rtsp stream from IP cameras, the dts/pts of the next frame occasionally will be smaller than the previous frames. Yet the doc of av_read_frame hasn't indicated whether any dts-out-of-order should be expected. Is it a bug of libav or a P frame or maybe there's a bug in the IP cameras' rtsp implementation? If that's expected, in order to to correctly handle it/mux it with av_interleave_write_frame, I have to cache and reorder those frames before write, do I? If this is not expected and it's probably a bug in the camera's rtsp codes, then I should discard the consequent frames with smaller dts when I mux it to avoid non increasing dts, should I?
Any guides/explanation or help are welcome. Thanks in advance. Get Outlook for Android<https://aka.ms/ghei36>
_______________________________________________ Libav-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/libav-user
