Raymond H. Dye - US grants

This award will help fund the creation of a 12-station, computer- controlled laboratory for the teaching of labs in Psychobiology, Hearing, Vision, Human Information Processing, Statistics, and Research Methods in the Department of Psychology at Loyola University of Chicago. The twelve stations will be networked so that software can be shared among the stations and so each can access the University's mainframe computer and the library's on- line data base. All courses offered in the facility emphasize "hand-on" approaches to learning, with students required to participate in the running of experiments, graphic presentation of the data, and data analysis. The proposal seeks to expand and supplement the existing facility, which has been used for Lab in Psychobiology, Lab in Vision, and Lab in Hearing. Software will be acquired that turns each of the twelve computer stations into a fully-functional laboratory, with four of the stations equipped with measurement, signal-generation, and physiological recording equipment necessary for the two sensory labs and Lab in Psychobiology. The grantee institution is matching this NSF award with funds from non-Federal sources.

This project deals with localization of multiple concurrently active sound sources. In order to localize the sources in such an environment the auditory system must correctly determine which spectral components arise from which location if the sources are to be correctly identified. The role of sound localization in this process is speculated to be great, since all components that comprise a particular source come from the same spatial location. The sound field at the two ears is said to be spectrally incoherent in that interaural differences of time and intensity, the two binaural cues for sound localization in the horizontal plane, vary across the spectrum. If binaural processing were to proceed as is suggested in the description above, it would be characterized as spectrally analytic in that the interaural differences of time associated with a particular spectral component allow it to be "heard out" as separate from other components that are present. Previous work has led to the conclusion that the binaural auditory system is often spectrally synthetic when the stimulus consists of a small set of low-frequency components presented over headphones -- the system integrates interaural information across the frequency domain without regard to which components give rise to which interaural differences. One example of binaural processing that is spectrally analytic is Huggins pitch. The existence of Huggins pitch is evidence that the binaural auditory system can, in some instances, segregate targets from a background based solely on differential interaural differences across the frequency domain. This project proposes a set of studies that attempt to further examine the extent to which listeners have access to information regarding which interaural cues are associated with which spectral components. Specific Aim 1 is an effort to ascertain the extent to which interaural information arising from different frequency components is independent. Experiments will measure the relationships between the ability of humans to detect interaural differences arising from different components of a complex and the ability to discriminate waveforms based on which component of the complex had been interaurally delayed. Specific Aim 2 is to develop and apply a variation of the "randomization technique of the general recognition theory" to further elucidate the manner in which interaural information arising from different spectral regions interacts. Specific Aim 3 is to study the extent to which manipulating the stimulus duration and the characteristics of the spectral and temporal fringe around the target affects listeners ability to analytically "hear out" targets from a background. this will be studied in stimulus configurations that lead to Huggins pitch and in conditions with small numbers of components. Specific Aim 4 is to examine the manner in which similarities and differences in the modulation of high-frequency components interact with interaural differences between targets and distractors to support segregation/fusion of targets with background components.

This Project, Project 3, Binaural Processing and Sound Source Determination, is a psychoacoustic study involving human listeners' use of spatial cues in sound source determination. The Project's focus is on stimulus conditions that involve locating more than one potential sound source. The Project continues work on the lateralization/localization of multiple spectral components using a newly developed two-dimensional stimulus classification procedure. This procedure estimates the relative weights listeners assign to various sound sources when they attempt to lateralize/localize one or more sources when all of the sound sources are presented at the same time. The cocktail party effect (the use of spatial cues in the ability to attend to one source when many sources are present) will be studied in a procedure that is similar to a real-work listening condition involving three or more concurrently presented sound sources. In many listening conditions the sound source and its echoes are all present when one attempts to locate the sources. A series of studies is planned to better understand how the auditory system processes echoes. These studies involve the Precedence effect, the Franssen effect, and the recently discovered Clifton effect. All of these effects indicate that spatial information concerning echoes is suppressed when a source and its echoes are localized. Recent results also suggest that a listener's immediate prior experience with a source and its echoes influences how echoes are processed. Studies are proposed to investigate these aspects of echo processing. Many stimuli can be generated which produce a 'dichotic pitch.' The generation of a dichotic pitch indicates that the binaural auditory system has used interaural differences to segregate one spectral region from another. Studies are planned investigating dichotic pitch stimuli as one means of understanding how interaural differences can be used to segregate one sound from another. Our proposed work, therefore, is based on a series of effects that appear to be related to the situation in which listeners are asked to determine the location of a sound source in the real world, especially considering that in the real work there is usually more than one sound source present. Thus, studies of binaural interference, the cocktail party effect, echo processing, and dichotic pitch all involve processing spatially separated sounds when more than sound source is present.