in reply to MP3 Concatenation

I would say the term you're looking for might be "production" or "post-production." Some of the production work can be automated, and some cannot.

MP3 format is a stream of chunks. I think for the most part, chunks are independent. There are modules that let you find the chunks and then you can order or redact them any way you like.

However, even if you decode each MP3 chunk to examine it as a waveform, it's not necessarily a computationally trivial task to decide if the chunk is "useful" or "chaff" which you can skip.

--
[ e d @ h a l l e y . c c ]

Replies are listed 'Best First'.
Re^2: MP3 Concatenation
by dimar (Curate) on Jun 11, 2004 at 17:10 UTC

    Hiya halley

    Some of the production work can be automated, and some cannot.
    Ok, one assumption was it would be necessary to first find the relevant 'useful' and 'chaff' sections by hand (eg by writing down the times, or by inserting a 'marker' inside the MP3 files). Of course this assumes there is a way to correlate the 'chunks' with time, or to add 'markers' to the chunks. The preliminary web searching seems to indicate that .wav files may support this, and that perl may not be a good way to go for this kind of production process. That would be a shame.

    Any additional insights welcome!

[reply]

      This is how it is done with the various opensource PVRs for the commercial cutting. The closed source (pay) Snapstream does it automatically with pretty good success

      I wonder if the start of the static can be detected (much like song start detection in many of the mp3 rippers) and marked for cutting.

      No one has seen what you have seen, and until that happens, we're all going to think that you're nuts. - Jack O'Neil, Stargate SG-1

[reply]

In Section Seekers of Perl Wisdom