The relationship between frequency selectivity and overshoot.

Under some conditions, threshold for a brief tone is higher at the onset of a broadband masker than it is if it is delayed from the onset of the masker. Evidence suggests that this "overshoot" is related to active processing in the auditory system. The present experiments examined this question, by measuring frequency selectivity under the same conditions in which overshoot was measured. The first experiment demonstrated that the growth of masking with masker level was approximately linear for a 1-kHz signal with or without a precursor (which was identical to the masker), and a 4-kHz signal with a precursor. For the 4-kHz signal with no precursor, an elevation in signal-to-masker ratio was seen at mid masker levels, relative to the other conditions. Frequency selectivity was then measured for a fixed-level signal, with and without a precursor. Relative frequency selectivity was highest for the 4-kHz signal with no precursor, lower for the 1-kHz signal with no precursor, and lowest for the 1- or 4-kHz signal with a precursor. The overshoot results and the frequency selectivity results would be consistent with stronger active processing at 4 kHz than at 1 kHz, and a decrease in active processing following a broadband noise precursor.