Blind Cancellation of Stereo-Base Widening

Cascading of stereo-base widening systems causes a deterioration of the listening performance. A solution to this problem is to cancel the widening effect before the signal is widened again. The main effect of stereo-base widening systems is to the effect on the phase difference. A model for the distribution of the phase difference of normal mixed stereo recordings is presented. Stereo recordings processed by a stereo-base widening system do not fit this model. An application of this property is illustrated with an example. This example can estimate the amount of widening of a particular widening system in an iterative blind way. At the same time it recovers the original audio signal.