Two-point discrimination in auditory displays

Abstract

In this paper we describe work which characterises the effect of spatial factors on the segregation of concurrent sound sources. The results inform the operational requirements of virtual auditory displays required to render multiple, concurrent sound sources in terms of (i) minimum spacing between sources and (ii) identification of the principal acoustic directional cues exploited by the auditory system for source segregation. Experiments using various broadband sound sources (white noise, click trains, spoken words) indicate that the extent of actual separation required for reliable segregation of concurrent stimuli varies as a function of location. The pattern of location dependence indicates that the auditory system is principally exploiting binaural differences for sound segregation. Monaural spectral cues, while essential for high fidelity spatialisation, seem to play a much less minor role in segregation under these conditions. However, spectral cues are likely to be useful when competing stimuli have distinct temporal structures or are not fully coincident in time.