Application-driven design of auralization systems

Acoustical environmental simulation (auralization) involves rendering a binaural signal that articulates simultaneously the positional information of a sound source and the source’s acoustic and vibratory interaction with its environmental context. This places significant challenges on both the modeling and rendering components of such a system. Overcoming real-time limitations of processors can be accomplished by implementing rendering limits based on auditory threshold data for both early reflections and late reverberant energy. However, final assessment of system quality is dependent on the specific task or goal of the simulation. The use of auralization in accomplishing a telerobotic task requires head tracking and low latency, while environmental cues such as obstacles with acoustic reflections can be simplified and exaggerated beyond veridical representation. Contrasting this is the task of assessing speech intelligibility of an emergency public address system, or the quality of a space for musical performance, requiring a more accurate level of simulation of acoustical materials and sound-source characteristics. Modeling accuracy is complicated by such factors as the level of variance in the absorptive and diffusive properties of materials in terms of their real-world application. Assessment of auralization quality must therefore involve a best estimate of significant factors for a particular application.