[FFmpeg-devel] [PATCH] examples/decoding_encoding: flush audio frames in the encoder

[Stefano Sabatini]

On date Wednesday 2012-08-22 01:26:13 +0200, Michael Niedermayer encoded:
> On Tue, Aug 21, 2012 at 10:32:01PM +0200, Stefano Sabatini wrote:
> > ---
> >  doc/examples/decoding_encoding.c |   14 ++++++++++++++
> >  1 files changed, 14 insertions(+), 0 deletions(-)
> 
> should be ok

Thanks, applied.

BTW, I wonder if this use of the API is correct:

        /* encode the samples */
        ret = avcodec_encode_audio2(c, &pkt, frame, &got_output);
        if (ret < 0) {
            fprintf(stderr, "error encoding audio frame\n");
            exit(1);
        }
        if (got_output) {
            fwrite(pkt.data, 1, pkt.size, f);
            av_free_packet(&pkt);
        }

What if we have more than one packet per frame?

Shouldn't we update pkt.data/size until *got_output is 0? This way we
avoid to cache frames in the encoder (especially if the frames
contains a huge amount of samples).
-- 
FFmpeg = Forgiving Fiendish Maxi Philosofic Erratic Game