Directional dependence of binaural loudness

Even though there is a vast body of literature on spatial hearing, little is known on how the location of a sound source affects its loudness. The pertinent research on head-related transfer functions(HRTFs) primarily deals with issues of localization and spatial discriminability. The literature on binaural loudness summation, on the other hand, focuses on a highly artificial paradigm according to which unnatural combinations of sound pressure levels(such as monotic-diotic comparisons) are presented to listeners via headphones. In order to reconcile these two approaches, we recently started a series of experiments determining the effect of sound incidence angle on loudness[1,2,3,4]. The basic experiment to be reported on here was conducted in an anechoic chamber, in which loudspeakers were mounted to present sounds from a number of directions both in the horizontal, and median planes. The subject’s task was to produce a loudness match to a frontal reference location, and matches were obtained via a two-interval adaptive forced-choice procedure. Narrow-band noises were used, in order to investigate the frequency-specific effects of direction on loudness. Furthermore, individual HRTFs were measured for each incidence angle, in order to relate the changes in actual at-ear exposure to the observed changes in loudness as a function of sound incidence. Combining the listening-test data with the physical measurements, an attempt was made to model the binaural summation underlying the directional loudness matches.