Now, leaving the language thing aside, let's discuss your scheme. If
you're doing audio compression you should answer the following points:
1) Is it lossy or lossless (clearly lossy from what you've written so far)
2) Does it require, or benefit from specialized hardware support?
3) How does it compare to existing compression standards on a
rate-distortion graph? How does it perform in subjective listening
tests?
4) What is your intended niche?
5) What kind of help do you think you require?
-- Dan