Effects of hair, clothing, and headgear on localization of three-dimensional sounds Part IIb

Abstract

Seven 20–25-year-old normal hearing (≤20 dBHL) native male-undergraduates listened twice to treatments of 85 virtual source locations in a large dark anechoic chamber. The 3-D-stimuli were anew-calculated white noise bursts, amplitude modulated (40-Hz sine), repeated after a pause (total duration 3×275=825 ms), HRTF-convolved and headphone-equalized (Sennheiser HD580). The HRTFs were measured from a Cortex dummy head wearing different garments: 1=alpaca pullover only; 2=1+curly pony-tailed thick-hair+eye-glasses; 3=1+long thin-hair (ear-covering); 4=1+mens trilby; 5=2+bicycle helmet+jacket [Riederer, J. Acoust. Soc. Am., this issue]. Perceived directions were signified by placing a tailored digitizer-stylus over an illuminated ball darkened after the response. Subjects did the experiments during three days, each consisting of a 2-h session of several randomized sets with multiple breaks. Azimuth and elevation errors were investigated separately in factorial within-subjects ANOVA showing strong dependence p(≤0.004) on all main effects and interactions (garment, elevation, azimuth). The grand mean errors were approximately 16°–19°. Confused angles were retained around the ±90°-interaural axis and cos(elev)-weighting was applied to azimuth errors. The total front−back/back−front confusion rate was 18.38% and up−down/down−up 12.21%. The confusions (except left−right/right−left, 2.07%) and reaction times depended strongly on azimuth (main effect) and garment (interaction). [Work supported by Graduate School of Electronics, Telecommunication and Automation.]