Title :
Using DSP-based parametric physical synthesis models to study human sound perception
Author :
Cook, Perry R. ; Lakatos, Stephen
Author_Institution :
CS Dept., Princeton Univ., NJ, USA
Abstract :
A series of studies in human auditory perception, memory, attention, discrimination, and learning are described, based on the use of physical synthesis models. The use of realistic sounding DSP-based physical models allows individual parameters to be isolated and tested, while still keeping an "ecological" approach to the experiments. Partly due to the flexibility of the model parameters, and due to the large numbers of subjects tested, and also due to the large collections of sounds we have been able to use, the studies described have provided much new evidence about the nature of human cognitive auditory mechanisms. A new research agenda has been born out of this work, endeavoring to answer difficult questions about the subjective nature of sonic "reality", "naturalness", "presence", "immersion", etc.
Keywords :
audio signal processing; hearing; sound reproduction; DSP-based physical models; human auditory attention; human auditory discrimination; human auditory learning; human auditory memory; human auditory perception; human cognitive mechanisms; human sound perception; parametric physical synthesis models; sonic immersion; sonic naturalness; sonic presence; sonic reality; Acoustic noise; Acoustic testing; Humans; Multi-stage noise shaping; Music; Noise level; Phase noise; Psychoacoustic models; Psychology; Shape control;
Conference_Titel :
Applications of Signal Processing to Audio and Acoustics, 2003 IEEE Workshop on.
Print_ISBN :
0-7803-7850-4
DOI :
10.1109/ASPAA.2003.1285823