[music-dsp] context-free gzip

David Olofson david at olofson.net
Tue Jan 28 02:41:00 EST 2003


On Tuesday 28 January 2003 10.41, Svante Stadler wrote:
[...]
> I think the main problem is that with these zip algorithms, 'almost
> the same' doesn't count. It has to be an exact repetition of a
> phrase or there will be no compression, correct? So the auditory
> model should be followed by some sort of dynamic vector
> quantization.

Actually, if you look closer at many compression algorithms, they use 
approximation/prediction in combination with a compact "error 
correction" data format. That is, find a low cost approximation 
that's close, and then encode the parameters for the approximation, 
as well as the difference between the original and the approximation.

That said, normal data compression algorithms are probably rather 
useless for audio anyway, since their approximation/prediction 
algorithms aren't looking for anything like audio. (Which is why they 
suck at compressing audio.)


//David Olofson - Programmer, Composer, Open Source Advocate

.- The Return of Audiality! --------------------------------.
| Free/Open Source Audio Engine for use in Games or Studio. |
| RT and off-line synth. Scripting. Sample accurate timing. |
`---------------------------> http://olofson.net/audiality -'
   --- http://olofson.net --- http://www.reologica.se ---




More information about the music-dsp mailing list