Jim J. Hant
Brian P. Strope
Abeer A. Alwan
Dept. of Elec. Eng., UCLA, 405 Hilgard Ave., Los Angeles, CA 90095
The reported perceptual experiments are aimed at quantifying the relationship between masked thresholds of signals within noise as a function of signal center frequency, duration, and signal type (pure tones or 1--8 critical-band noises). Tests are adaptive 2AFC with the signals' center frequencies ranging between 0.4 and 6 kHz and their durations, between 10 and 300 ms. The masker is flat noise with a spectrum level of 36 dB SPL/Hz sampled at 16 kHz. Four subjects participated in the experiments. Critical-band theory accurately predicts the masked thresholds for tones at 300 ms. Tone thresholds increase almost linearly with decreasing logarithmic duration, and these durational effects are more pronounced at low frequencies. Thresholds for 1 CB noise and tone signals are similar at short durations (10--30 ms). For longer durations and for wideband noises, however, differences both in thresholds and the way the thresholds change with duration are observed. Possible explanations in terms of changing auditory filter bandwidths, temporal integration, temporal onset detection and/or an interplay between these factors will be presented. The data are used to predict the masked thresholds of stop bursts in the presence of noise. [Work supported by NIH.]