Relationship between Psychophysical Tuning Curve and Critical Band

A two-interval, two-alternative forced choise task was used to estimate psychophysical tuning curves (PTCs) in simultaneous masking. In each experiment, the masked PTCs for signal levels of 10, 20 and 30dBSL caused by the addition of a 500ms masker were determined as a function of masker frequency. The results were as follows: 1) The PTC for the highest signal level had the shallowest lower branch and steepest upper branch, and it had the narrowest tip. The PTC for the lowest signal level had the steepest lower branch and shallowest upper branch, and it the broadest tip. These properties of PTCs had been explained in terms of excitation-pattern model. 2) The bandwidths of PTCs were not dependent on the level and durations of signal, however, if we measured the bandwidths of tuning curves and filter shapes 10dB from the tip, we found that the widths of the auditory filters are 1.5-2.0 times the widths of the PTCs. This phenomenon was adequate to explain the basic effects of off-frequency listening on PTCs. 3) Temporal integration for signal level of 10 and 20dBSL were shown, however, for its level of 30dBSL, PTC tips were independent of signal duration. From these results, we found that the masked threshold of short duration and high level signal was independent of critical band. 4) The conventional PTC associated with combination tone and off-frequency gave a more accurate indication of frequency selectivity of the auditory system and showed that the bandwidth of PTC was to reflect properly the critical bandwidth.