Application of digital signal processing techniques to research on sound localization

Abstract

The availability of fast, general‐purpose laboratory computers has dramatically changed the way research is done in many areas of psychoacoustics. In studies of sound localization, for example, standard digital signal‐processing techniques are used to synthesize stimuli for headphone presentation that contain all of the localization cues (e.g., interaural differences, pinna effects) that are available from sounds presented in free field. This brings to localization research a much needed degree of stimulus control and specificity, and allows us to address certain basic issues that were heretofore inaccessible. This presentation will survey the general principles involved in the application of digital signal processing to localization research, and will discuss the practical limitations of the specific techniques that are used (e.g., FFTs, FIR filters, inverse filtering). Empirical data will be shown in order to address questions relating to the classic tradeoffs among signal/noise ratio, bandwidth, word‐length, and sampling frequency. The feasibility of real‐time digital processing in currently available PC‐based systems will also be discussed. [Work supported by NIH, NSF, and NASA.]