Three-dimensional sound synthesis based on head-related transfer functions

Abstract

A common technique for synthesizing three-dimensional sound makes use of head-related transfer functions (HRTFs). The HRTFs made available to the public by Gardner and Martin of MIT Media Lab. are used with the intent to refine them so as to exaggerate the perceptual differences for sounds coming from different directions, and to emphasize the pinna effects. With the refined HRTFs reduced front-back reversal rates and improved ability to locate moving sources are demonstrated. * Manuscript received 1996 October 21; revised 1998 February 18 and July 20.