The environment often contains a great deal of extraneous sound (noise) from which listeners nevertheless can sometimes extract information from particular sounds (signals). The mechanisms by which this is accomplished are still not well understood. The importance of clarifying them is not only the basic scientific interest and in basic mechanisms working in auditory perception. In many instances the extraneous sounds prevent the listener from extracting information from the signal. Knowing how to minimize this problem is important in designing systems for auditory communication in many applications. In this project the investigator is attempting to test his hypothesis concerning this problem, which is that in listening we compare sounds of various frequency regions (pitch) and different intervals of time, weight them in some way, and separate the signal from the noise in analyzing the result. He will present subjects with reproducible noise samples in which the frequency spectral and temporal fringes are systematically varied. Using computers he will stimulate various models of excitation, inhibition and other interactions to determine which models (if any) accurately account for the effects observed by the listeners.
