On Feb 25, 2013, at 2:29 AM, Carl Eugen Hoyos <ceho...@ag.or.at> wrote:
> Before you start debugging (the cast to sourceData looks > suspicious): Did you look at doc/examples/filtering_audio.c > and doc/examples/resampling_audio.c ? > I suspect using the aconvert filter has the advantage that > you can do other changes to the audio without additional > code (and bugs). My source code is modeled directly after resampling_audio.c. After our prior discussions about merits of each, history of Libav, etc. it seemed clear that libswresample was the proper direction, and that resampling_audio.c recommended the generally recognized conventional approach. Are there known bugs or gotchas in libswresample that may be possibly hit by using the resampling_audio.c approach? Before scrapping any resampling_audio.c-modeled approach (given that everything except the last line works, and there's no guarantee taking the aconvert route won't end up at the same obstacle), I'd like to understand what reason that this approach won't suffice...that will be useful information to know. Thanks, Brad _______________________________________________ Libav-user mailing list Libav-user@ffmpeg.org http://ffmpeg.org/mailman/listinfo/libav-user
