avconv: Don't split out inline sequence headers when stream copying VC1
authorMartin Storsjö <martin@martin.st>
Mon, 23 Jan 2012 11:06:07 +0000 (13:06 +0200)
committerMartin Storsjö <martin@martin.st>
Tue, 14 Feb 2012 14:12:15 +0000 (16:12 +0200)
This is required when stream copying VC1 in ismv - there's one
global header in the moov atom, but keyframes have a separate
sequence header prepended.

Signed-off-by: Martin Storsjö <martin@martin.st>
avconv.c

index 92ffd0fd0d5c6889f69d1501e7ed9f9986324505..862e325ad96222083ce0093ecde9de4ed0fb06aa 100644 (file)
--- a/avconv.c
+++ b/avconv.c
@@ -1795,6 +1795,7 @@ static void do_streamcopy(InputStream *ist, OutputStream *ost, const AVPacket *p
     if (  ost->st->codec->codec_id != CODEC_ID_H264
        && ost->st->codec->codec_id != CODEC_ID_MPEG1VIDEO
        && ost->st->codec->codec_id != CODEC_ID_MPEG2VIDEO
+       && ost->st->codec->codec_id != CODEC_ID_VC1
        ) {
         if (av_parser_change(ist->st->parser, ost->st->codec, &opkt.data, &opkt.size, pkt->data, pkt->size, pkt->flags & AV_PKT_FLAG_KEY))
             opkt.destruct = av_destruct_packet;