Spatial acoustic cues for the auditory perception of speaker's facing direction

In pursuit of an ultimately realistic human-to-human telecommunication technology, the ability to auditorily perceive the facing direction of a human speaker was explored. Listeners’ performance was assessed in an anechoic chamber. A male speaker sat on a pivot chair and spoke a short sentence while facing a direction that was randomly chosen from eight azimuthal angles or three elevation angles. Twelve blindfolded listeners heard the spoken sentence at a distance of either 1.2 or 2.4 m from the speaker and were asked to indicate the speaker’s facing direction. In separate sessions, the speaker continuously changed facing angles while speaking and the listeners indicated the perceived direction of horizontal rotation (clockwise or counter-clockwise) or vertical rotation (upward or downward). The overall results showed that the listeners’ average response errors were 23.5 degrees for azimuth and 12.9 degrees for elevation. These values were comparable to or better than those obtained in previous studies using a loudspeaker. The average correctresponse rates for rotation direction (either horizontal or vertical) were equal to or more than 80%. To identify acoustic cues that have caused the listeners’ accurate performance, the acoustic transfer characteristics from the speaker’s mouth to the listener’s ears were measured by the cross-spectral method. Finer transfer functions were further obtained in a couple of conditions of particular interest by numerical computer simulation using the finite difference time-domain method. The results suggested that major cues included but were not limited to the overall level and spectral tilt for the front-back or up-down judgment, and the interaural level difference for the left-right judgment.