Using an excitation-pattern model to predict auditory masking

This paper evaluates the extent to which auditory masking can be reliably predicted from excitation patterns. For this purpose a quantitative model proposed by Glasberg and Moore [Hear. Res. 47, 103-138 (1990)] was used to calculate excitation patterns evoked by stationary sounds. Model simulations were performed for a number of masking experiments, reported in the literature, by calculating excitation patterns for the masker-alone as well as for the masker-plus-target conditions. As a threshold criterion, a difference between the two patterns of 1 dB at any frequency was imposed. For narrow-band-noise masking patterns, the method yields a fairly precise prediction of experimental data. For other conditions, however, systematic deviations between model predictions and data are observed. For instance, the model does not reproduce the typical tip-tail shape of psychophysical tuning curves. Furthermore, the nonlinearities in level dependence are not correctly described, and the model fails to reproduce a realistic two-tone masking curve.