Method of generating 3 dimensional sounds for elevation perception

Abstract

The performance of various sound localization cues, those are stationary sound localization cues and dynamic sound localization cues, in providing elevation impression of virtual sources displayed along the median plane was compared. Among different dynamic sound localization cues, a dynamic cue from up and down head-motion (tipping motion) was the most frequently observed when sound sources were located at the frontal median plane. Listeners’ sound source localization ability was significantly improved by allowing their tipping head-motion both for actual and virtual sources located at the median plane. In reproduction of virtual elevation impression along the median plane, it should be required to consider listeners’ tipping head-motion because listeners can easily recognize that the perceived impression is not elevated but is at the actual loud speakers’ elevation position that perception is based on his/her tipping head-motion during listening sounds. One way of considering tipping head-motion is selecting configuration for loud speaker system that can minimize resultant dynamic cue induced from listener’s tipping head-motion. A stationary sound localization cue can provide virtual elevation impression to listeners only after a resultant dynamic cue for a loud speaker system is minimized. For stationary sound localization cues, sound reflection from our shoulder and torso is of minor importance. Only pinnae and head response can provide a similar or better sound localization performance compared to when both pinna and head response and shoulder and torso response were provided.