Hello ffmpeg-devel! These two patches aim to improve playback support for Audible Format 2/3/4 as implemented in libavformat/aadec.c.
Patch 1 is a very simple improvement of EOF detection in the presence of trailing data and broken/truncated files. It consists mainly of some extra error/bounds checks. Patch 2 adds chapter support and format-specific seeking; due to the chunk/chapter nature of the content, the two features are very much intertwined and cannot easily be split into separate patches. The generic index seek was enabled before, but usually hit the middle of a content chunk and thus caused garbled playback. The patch fixes this by ensuring proper chunk alignment and read state update. Since I am new to ffmpeg development, I am not quite sure if my solution always does things the way they are supposed to be done (aka best practice). In particular, I am setting the stream time_base according to the content bitrate in bytes-per-second, which makes for very convenient conversion between timestamps and file positions (when removing/adding the chapter headers from the equation). While it works nicely, I cannot tell if this breaks things somewhere else. Please let me know what you think, improvements very welcome! Disclaimer: No reverse engineering was involved in creating this patch. Everything was derived from preexisting code, mostly aa_read_packet. Cheers, Karsten Karsten Otto (2): aadec: improve eof detection aadec: add chapters and seeking libavformat/aadec.c | 86 +++++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 81 insertions(+), 5 deletions(-) -- 2.14.3 (Apple Git-98) _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
