On date Sunday 2016-06-05 12:59:48 +0200, Stefano Sabatini encoded: > On date Thursday 2016-05-19 18:50:17 +0200, Stefano Sabatini encoded: > > On date Thursday 2016-05-19 18:45:22 +0200, Stefano Sabatini encoded: > > > This format is useful to inject custom user data into streams. > > > --- > > > doc/demuxers.texi | 40 +++++++++ > > > doc/muxers.texi | 31 +++++++ > > > libavformat/Makefile | 2 + > > > libavformat/allformats.c | 1 + > > > libavformat/fftextdatadec.c | 212 > > > ++++++++++++++++++++++++++++++++++++++++++++ > > > libavformat/fftextdataenc.c | 103 +++++++++++++++++++++ > > > 6 files changed, 389 insertions(+) > > > create mode 100644 libavformat/fftextdatadec.c > > > create mode 100644 libavformat/fftextdataenc.c > > > > Short explanation: I needed some way to inject serialized data packets > > through stream-copy, so I implemented this format. The other patches > > are related (since I needed to inject timed ID3 data). > > > > I'm fine with extending it with other options, or to use an > > alternative solution (if exists). > > Please tell if you consider this format acceptable. > > One obvious limitation is that it can't store codec properties (nor > the codec name, nor more than one stream). One possibility would be to > dump the serialization of AVCodecContext/AVCodecParameters, even if > these would require a serious limitation on the format (since that > means it would require to use it only as intermediary format, for > muxing and demuxing using the same version of FFmpeg, which is fine in > my use case).
Ping. Any hope to get this integrated, or there is someone against it? -- FFmpeg = Fiendish & Faithful Magnificient Pure Everlasting Gladiator _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
