[FFmpeg-user] conversion from h.264 to mp3 adds 0.12 seconds of silence to beginning of file

[Reid Oda]

Hi list,

I'm doing audio/video analysis on videos downloaded from youtube. As part
of the process, I need to extract the audio to a form that can be opened by
my audio analysis software. I've been using mp3, but I recently discovered
that ffmpeg adds 0.12 seconds to the beginning of the .mp3 file. This
causes sync problems.

Here is the info on the youtube videos:

Stream #0:1(und): Audio: aac (mp4a / 0x6134706D), 44100 Hz, stereo, fltp,
96 kb/s (default)

Any thoughts on how to make sure that the extracted audio is exactly in
line with the original aac audio (unfortunately, my software can't read
aac)? Right now I am simply adding the offset onto my computations, but I
would prefer to have a reliable conversion.

Thank you very much for any help!