I'm using the FFmpeg SDK to programmatically convert videos to mp3.
I read the audio frames of the video this way:
while(av_read_frame(pFromCtx, &pkt) >= 0)
{
if(pkt.stream_index == audioStreamIndex)
{
avcodec_get_frame_defaults(frame);
got_frame = 0;
ret = avcodec_decode_audio4(pCodecCtx, frame, &got_frame, &pkt);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error decoding audio frame.\n");
continue;
}
if(got_frame)
{
// Write the decoded audio frame
write_audio_frame(pToCtx, pToCtx->streams[pToCtx->nb_streams-1], frame);
}
}
av_free_packet(&pkt);
}
Decoding the audio from the input video file works fine. The problem occurs when I try to encode a mp3 frame:
static void write_audio_frame(AVFormatContext *oc, AVStream *st, AVFrame *frame)
{
AVCodecContext *enc = st->codec;
AVPacket pkt;
int got_packet = 0;
int ret = 0;
av_init_packet(&pkt);
pkt.data = NULL;
pkt.size = 0;
ret = avcodec_encode_audio2(enc, &pkt, frame, &got_packet);
if (ret < 0) {
// PROBLEM
fprintf(stderr, "Error encoding audio frame. \n");
exit(1);
}
}
I get the following console output:
[libmp3lame] inadequate AVFrame plane padding
The only happens with .flv files, the code works fine for .mp4 files. Any clue what the error message means?
Thanks