[flac-dev] FLAC Mathematical Details

[Ralph Giles]

On 7 February 2012 06:24, Manu Ghulyani <manughulyani at gmail.com> wrote:

> I am unable to find the mathematical details of the codec , please guide.

Unfortunately I'm not aware of a clear write-up of how the FLAC format
works in detail. One must muddle through between
http://flac.sourceforge.net/format.html and the source code.

Basically the audio is chopped into a blocks and each block is coded
either uncompressed, as a constant value (good for silence), or with
linear predictive coding plus a rice-coded residual. I don't know how
the encoder decides where to put the block boundaries.

Hope that's enough to get you started,
 -r