# HG changeset patch # User michael # Date 1235877444 0 # Node ID 108bb305ba6f40450ea3f68355058e3bc961160f # Parent 5b9eddbee9c4bf7a7ae08c276040dc201c2d5d56 Disable MPEG-1/2 style timestamp calculation for H264. It still randomizes the timestamps because delay is not known for the first few frames. diff -r 5b9eddbee9c4 -r 108bb305ba6f utils.c --- a/utils.c Sun Mar 01 00:12:08 2009 +0000 +++ b/utils.c Sun Mar 01 03:17:24 2009 +0000 @@ -859,7 +859,8 @@ // av_log(NULL, AV_LOG_DEBUG, "IN delayed:%d pts:%"PRId64", dts:%"PRId64" cur_dts:%"PRId64" st:%d pc:%p\n", presentation_delayed, pkt->pts, pkt->dts, st->cur_dts, pkt->stream_index, pc); /* interpolate PTS and DTS if they are not present */ - if(delay==0 || (delay==1 && pc)){ + //We skip H264 currently because delay and has_b_frames are not reliably set + if((delay==0 || (delay==1 && pc)) && st->codec->codec_id != CODEC_ID_H264){ if (presentation_delayed) { /* DTS = decompression timestamp */ /* PTS = presentation timestamp */ @@ -907,7 +908,7 @@ FFSWAP(int64_t, st->pts_buffer[i], st->pts_buffer[i+1]); if(pkt->dts == AV_NOPTS_VALUE) pkt->dts= st->pts_buffer[0]; - if(delay>1){ + if(st->codec->codec_id == CODEC_ID_H264){ //we skiped it above so we try here update_initial_timestamps(s, pkt->stream_index, pkt->dts, pkt->pts); // this should happen on the first packet } if(pkt->dts > st->cur_dts)