Azimuthal tuning of human perceptual channels for sound location.

Human sound localization is acute for frontal locations, but relatively poor in the lateral hemifields. Previous studies in man have not, however, provided evidence on the tuning of the perceptual channels for auditory space that subserve this pattern of acuity. The spatial tuning of perceptual channels used in human azimuthal sound localization was determined using a between-channel auditory temporal gap detection paradigm. In this paradigm, gap thresholds are low when the markers bounding the silent period (gap) activate the same perceptual channel but are elevated when the two markers activate different channels. To determine the tuning of spatial channels, gap thresholds were obtained in an anechoic room with white noise markers coming from each combination of 12 leading marker locations and 18 trailing marker locations throughout the full 360 degrees of azimuth in the horizontal plane through the interaural axis. Gap thresholds remained low (2-4 ms) for all combinations of leading and trailing markers between 30 degrees and 150 degrees in both lateral hemifields. When the leading marker was located deep in one hemifield, and the trailing marker was in the opposite hemifield, gap thresholds rose to 8-16 ms. For leading marker locations at 30 degrees from the midline, gap thresholds were low for all trailing marker locations in the ipsilateral hemifield and locations near the midline in the contralateral hemifield, and were elevated (6-8 ms) only near the contralateral pole. Finally, for leading marker locations at 0 degree or 180 degrees, gap thresholds were low for any trailing location within 30 degrees of the midline at the front or back, and thresholds were elevated for trailing locations at the lateral poles. These data are accountable in terms of two broadly tuned perceptual channels, occupying the left and right auditory hemifields, respectively, each extending 30 degrees across the midline. These channels have widths and locations similar to the spatial receptive fields previously described for central auditory neurons in animals. The data suggest a model of spatial acuity based on the rates of activation of two spatially overlapping channels, rather than the selective activation of members of a large population of finely tuned channels.