There are usually several sources of sound that we hear, all at the same time, yet the auditory system is quite good at segregating which sound components come from which source. A sound source which is localized and identified based on its acoustic properties has been called an "auditory image" and may depend on several features such as the frequency, intensity, and pattern of its acoustic spectrum. This project investigates the role that coherent amplitude modulation (AM) may play in the formation and segregation of auditory images. Processing of AM is evaluated with psychophysical detection, discrimination, and masking procedures, to examine both how information on the modulation is extracted, and how this information is combined across frequency regions to aid auditory image formation. Specific models will be tested along with computer simulations of subject performance. The experiments use novel approaches to important questions of complex auditory functions, and results from this basic work also are likely to have some impact on automated speech recognition systems and aids for the hearing impaired.
