Sivonen, Ville Pekka ; Ellermeier, Wolfgang (2007)
A model for predicting binaural loudness of spatial sounds.
30th annual midwinter research meeting of the Association for Research in Otolaryngology. Denver, Colorado, USA (10.02.2007-15.02.2007)
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

The location of a sound source relative to the listener greatly affects the physics of binaural stimulation: Depending on the angle of incidence, the transfer of sound from the source to the ears changes dramatically. As a consequence of these dependencies, which can be described by head-related transfer functions (HRTFs), the sound signals reaching the two ears often differ from one another, both in terms of their time of arrival and their spectral characteristics. In order to predict the loudness of spatial sounds, it is thus important to know how the two, often different signals reaching the listener's ears are integrated into a single binaural percept. Current loudness models [e.g., Moore et al., J. Audio Eng. Soc., 45, 224-239, (1997)] assume "perfect summation" of loudness across the ears, that is, binaural loudness being a simple sum of the monaural loudnesses in sone units. This assumption, however, is based on headphone studies quantifying the binaural effect on loudness, in which the effects of HRTFs and the spatial aspects are being disregarded. The present paper summarizes results from a series of experiments reporting on (1) how the spatial location of a sound affects its perceived loudness, and (2) how the binaural summation of the at-ear signals should be modeled to predict the loudness of sounds from various locations. The results provide evidence for less-thanperfect summation of loudness when truly spatial sound stimulation is used. More specifically, a power summation of at-ear spectra makes fairly accurate predictions of the mean data obtained.

Frage zum Eintrag

Redaktionelle Details anzeigen