Auditory segregation of concurrent signals: an operational definition

Auditory segregation of simultaneous, multiple (quasi-) steady-state acoustic signals which compose an ensemble implies simultaneous resolution along three "cardinal dimensions": vectors of pitch (residue plus timbre), temporal structure, and location. It is proposed that the normalized distance between two, or several, component signals in an ensemble along any of the three dimensions is traded off by differences on the two other dimensions. A psychophysical paradigm is described for objective measurement of auditory segregation and a metric is suggested for the prediction of segregation in the 3D parameter space.