Combining Spatial and Non-Spatial Sonification for Sound Localization in Virtual Environment: a Subjective Study

Abstract

Spatially sonified audio is frequently used in virtual environments to serve as an object localization cue. However, spatializers that are available to the general public typically have their limitations and lead to sound localization errors. By combining it with non-spatial sonification, the technique of encoding information with perceptual attributes of the sound, the localization errors of spatially sonified audio could be reduced. However, the encoded information would require higher mental effort to understand, which could be a concern under high task load. In this paper, we conducted a subjective study to compare spatial sonification against combined auditory cue using spatial and non-spatial sonification for target localization. We tested both methods under low and high task load. Our results show that combined sonification could significantly reduce the errors in localizaiton, but the advantage would be influenced by higher task load. Further, we compared several different non-spatial sonification methods for encoding depth information. Our experiment shows that both pitch and pulse frequency modulation are feasible for presenting depth of the target. Our results could contribute as a first step to a general guideline of designing sonification method as target localization cue.