Workshop on the Auditory Basis of Speech Perception

Keele University, UK
July 15-19, 1996

Implementation of a Psychoacoustical Preprocessing Model for Sound Quality Measurement

Martin Hansen, Birger Kollmeier

AG Medizinische Physik, Carl von Ossietzky-Universität Oldenburg, Germany

This study investigates the applicability of a psychoacoustical preprocessing model for measuring the speech quality of low-bit-rate codecs. The method for deriving an objective measure of the sound quality is to transform the input and output signal of a speech coding device into a so-called "internal representation" of the sound. Differences between the internal representations of input and output signal, calculated as the correlation coefficient between them, correspond to a decreased quality of the output signal. The preprocessing model consists of a critical-band filterbank which simulates the filtering by the basilar membrane, followed by a nonlinear adaptation circuit modeling amplitude compression and temporal masking effects, and a further filtering stage which analyzes fluctuations of the temporal envelope of the sound signal. This signal processing seems to be a successful "universal" model to simulate a variety of psychoacoustical data as well as the perception of speech sound at the same time.

Full Paper

Bibliographic reference.  Hansen, Martin / Kollmeier, Birger (1996): "Implementation of a psychoacoustical preprocessing model for sound quality measurement", In ABSP-1996, 79-82.