Hello All,

I've been searching the archives (unsuccessfully) for tips/code to do  
Blind [Signal, Source] [Separation, Extraction].
Particularly [Independent, Prior] Subspace Analysis but not necessarily.

The idea is to take a single audio input and identify, at any point in  
time, the number of original sources. They may or may not have  
harmonic spectra and can vary in pitch; a sort of primitive polyphonic  
line extraction, known as Blind Source Separation.

I know some of the theory but I'd like to some examples. Anyone?

thanks,


David


________________________________________________
Dr David Worrall
- Experimental Polymedia:	www.avatar.com.au
- Sonic Communications Research Group:	creative.canberra.edu.au/scrg
- MARCS Auditory Laboratories: marcs.uws.edu.au