Spectral precedence methodologies in spatial auditory research: perceptual evaluation of spatial attributes in 2D and 3D virtual auditory display environments

Sounds are space specific. That is, the spectral content of a sound suggests a perception that resides along the three co-ordinates of physical space; front to back, left to right and up to down. Each of these corresponds respectively to the spatial parameters source distance/depth, source width/spatial extent and volumetric height impression. Historically, spectral diffusion as a method in spatial audio has been the domain of electroacoustic practitioners who have argued for the promotion of space in artistic considerations. In parallel, engineering research in the field of spatial audio has investigated time frequency spatial methods as a way to implement signal decorrelation to increase the apparent size of a source in the horizontal plane. This thesis provides an in-depth review of existing literature in the fields of spatial audio development that spans audio signal processing, auditory cognition, panning algorithms, spatial audio applications and psychophysical testing methods. The research examines spectral diffusion from the perspective of expanding upon its current implementation method including; volumetric imaging in the horizontal and vertical planes, distance imaging for proximity (inside) and depth (outside) of an array and phantom imaging is successfully applied to 2D and 3D arrays. It is proposed that the method improves upon existing source placement panning approaches and provides much potential for the development of future spatial audio applications.