L'octidi 28 messidor, an CCXXII, Michael Niedermayer a écrit : > again, i suspect this is not updated by filters currently
You are right, the filters do not update them currently. I believe changing that policy would not be a convenient way of solving that issue: the duration information raises consistency problems: what happens if next_frame->pts is not equal to cur_frame->pts + cur_frame->duration. Also, that would be abusing a field, since pkt_duration is the duration of a packet, not a frame. I believe a better solution would be to send EOF pseudo-frames forward in the filter chain instead of relying on the backward request_frame() calls to return EOF: it would solve this particular issue (a pseudo-frame has a timestamp) and other issues as well. This is one of the main things I intend to look at during this summer. Regards, -- Nicolas George
signature.asc
Description: Digital signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
