On Sun, 21 Sep 2014 15:59:43 +0200 James Darnley <james.darn...@gmail.com> wrote:
> On 2014-09-21 12:51, wm4 wrote: > > it's easy to prevent at least early EOF. Playback will happen to work, > > even if it's slightly broken. > > Do we want to print a warning in the "slightly broken" case? Does it > just mean that the middle cannot be gapless between file1 and file2? > Well, it means that while decoding the packet stream, the gapless info of file2 will be applied. If file1 is longer than file2, some random samples in the middle of file1 will be dropped. If file2 is longer, some random samples in the middle of file2 will be dropped. One could just print a warning if mp3 data is successfully parsed after the expected end (end according to the length of the file in frames stored in the gapless info), but I didn't want to add another flag to the context. (Without such a flag you would print the warning after every packet read.) Another possibility: guess whether the stored length in frames doesn't match up with the byte size and the VBR index. But sounds hard and not very reliable. Or the whole file could be parsed. A libavformat mode that fully parses and indexes a file on user request would probably be nice to handle all these raw audio formats. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
