On Tue, Nov 24, 2015 at 03:23:17AM -0600, Rodger Combs wrote: > This caused issues when seeking in some unusual MPEGTS files > --- > libavcodec/aac_ac3_parser.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/libavcodec/aac_ac3_parser.c b/libavcodec/aac_ac3_parser.c > index 2f7d568..b01d5a5 100644 > --- a/libavcodec/aac_ac3_parser.c > +++ b/libavcodec/aac_ac3_parser.c > @@ -34,6 +34,7 @@ int ff_aac_ac3_parse(AVCodecParserContext *s1, > ParseContext *pc = &s->pc; > int len, i; > int new_frame_start; > + int got_frame = 0; > > get_next: > i=END_NOT_FOUND; > @@ -51,6 +52,7 @@ get_next: > if(len<=0){ > i=END_NOT_FOUND; > }else{ > + got_frame = 1; > s->state=0; > i-= s->header_size -1; > s->remaining_size = len; > @@ -80,7 +82,7 @@ get_next: > and total number of samples found in an AAC ADTS header are not > reliable. Bit rate is still accurate because the total frame duration > in > seconds is still correct (as is the number of bits in the frame). */ > - if (avctx->codec_id != AV_CODEC_ID_AAC) { > + if (avctx->codec_id != AV_CODEC_ID_AAC && got_frame) {
is this not also needed for bitrate ? and patch should be ok [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB I am the wisest man alive, for I know one thing, and that is that I know nothing. -- Socrates
signature.asc
Description: Digital signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel