I know it'd be easy to implement in DSP, the challenge is to get the micro to do it. Everything i've been taught has gone completely against this, so it was a suprise when i heard it could be done.
Is this a statistics issue or is this an algorithm problem, any page where i could get some info on it?