The method of claim 14 wherein the input signal is an audio signal, each sample represents air pressure at a given point of time, the first component signal represents magnitude changes in air pressure, and the second component signal represents shifts in time.