On 5/10/16, Carl Eugen Hoyos <ceho...@ag.or.at> wrote: > Paul B Mahol <onemda <at> gmail.com> writes: > >> OT but I remembered seeing overread messages from OP >> when using your patch on his files. > > How is the fact that the device produces damaged files > (be it because of wireless or other issues) related to > fixing broken audio for the given samples?
Either too big predictors should be skipped or clipped. How decoded output sounds when played back when predictors are clipped? Note that both approaches are flawed as your patch, if bytes written in damaged files are small numbers... _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
