FPGA Central

Hi. Just out of curiosity, has anyone designed and built their own
method for music compression. I realise that the ogg format is
open-source and relatively unencumbered by licencing issues. But I'm
curious if anyone has written their own compression method "because
they can".

Cheers,

Ross-c

I have not implemented such a compressor, but:
I would like to know if there are any signature to distinguish Music
from random noises?
Amos

I'm sure I saw a paper in the JAES some years ago which classified
music according to genre. Though I'm not sure I was very impressed by
the technology used, I think it was a rule based classifier. I don't
have the reference.

One problem is that one person's random noise is another person's
music. I would think that it would be fairly simple to distinguish
music with a regular beat from random noise, but what happens when you
put Frank Zappa's _Lumpy Gravy_ through it?

Cheers,

Ross-c

Ross Clement (Email address invalid - do not use) wrote:
> Hi. Just out of curiosity, has anyone designed and built their own
> method for music compression. I realise that the ogg format is
> open-source and relatively unencumbered by licencing issues. But I'm
> curious if anyone has written their own compression method "because
> they can".
>

I tried to do something like that just to see how redundant is the music
information, and if there is a long term correlation in the music. My
goal was to do the loseless compression with variable bitrate. Basically
what it was backward LPC analysis + Huffman postcompression.

Vladimir Vassilevsky

DSP and Mixed-Up Signal Design Consultant

http://www.abvolt.com

I'm doing something similar to this now, except I hope to try a few
"additions" to it. I would eventually have a go at sub-band coding, but
not soon. Since I use Linux both at home and work, if I made my own
compression method I could easily use it as my default.

Cheers,

Ross-c

Hi Ross,

maybe the Monkey's Audio packer would be interesting to look at:
http://www.monkeysaudio.com/developers.html


Martin

--
Quidquid latine scriptum sit, altum viditur.

Thanks. I must point out that my aim is not to design an audio
compression method to use, my aim is to learn more about the underlying
technology, and any actual compressor I build will just be a
side-effect of that. But the monkey audio packer (providing I can get
the linux port working) will give me a baseline to compare my own
compression to.

Cheers,

Ross-c