Directional hearing under progressive impoverishment of binaural cues.

Auditory stimuli--bursts of broadband, 4.0-kHz high-pass and 1.0-kHz low-pass noise--generated by loudspeakers arrayed in the horizontal plane were picked up by miniature microphones placed in the external ear canals and recorded on tape. When these recorded sounds were presented through headphones, listeners were able to identify the various loudspeakers that originally generated the sounds about as accurately as they could when the sounds were presented free-field. When the recorded sounds were processed so that either one or more of the interaural difference cues were removed directly or their removal was approximated, performance proficiency depended upon the frequency composition of the stimulus. It appeared that accuracy in identifying the loudspeaker that originally generated the broadband and 1.0-kHz low-pass noises could be maintained by the presence of interaural temporal differences. Accuracy on the task for the 4.0-kHz high-pass noise depended in large part on the presence of interaural intensity differences. No one performed proficiently when only interaural spectral differences were available.