Sound Localization

Abstract

When one listens to a transient type signal from two equidistant loudspeakers in free space, he experiences the sensation of a single well-defined source. The apparent position of the source can be altered by adjusting the relative outputs of the two speakers, by delaying the signal from one or by a combination of level and delay adjustments. If the transient type signal is replaced by a steady-state single frequency tone, the virtual source is again well-defined and can be changed in azimuth position by altering the speaker outputs as long as the two sources are in phase. However, as one source is delayed the virtual source becomes indistinct and localization becomes difficult. Tests will be described which indicate that when one listens to a steady-state single frequency tone, as from a pair of sources in free space or from a single source and its images in a room, he experiences the sensation of a well-defined single source only when the resultant time and pressure differentials correspond to those which could result from a single source in free space. This indicates that the apparent position of a steady-state single frequency source in a room bears a rather complex relation to the position of the actual source. Correct localization is dependent on transients in the signal and may be influenced by visual cues.