Rendering moving sound source for headphone-based virtual acoustic reality aspects of signal processing implementation

Abstract

This paper focuses on aspects of implementation of moving virtual sound source in virtual auditory environment. Rendering of dynamic source based on use of Head-Related Transfer Function is achieved by application of filtering with varying impulse response of the filter. This algorithm may be in its essence implemented by several methods, where each approach provides slightly different output signal resulting in variance of perception of spatial fidelity. The main difference results from different output of linear and circular convolution usually implemeted in such processing. Particular variants are described, discussed and compared by subjective listening test of the positioned signals. Other aspects of implementation as artifact occurrence are also discussed.