Critical Bands and Two-Tone Suppression in Band-Limiting Experiments

Abstract

Band-limiting experiments were performed at six test signal frequency (0.5, 1, 2, 4, 6 and 8kHz) for simultaneous, forward and backward masking. Masked threshold for a masker noise spectrum level 30dB SPL was examined as a function of the bandwidths of the bandpass, lowpass and highpass masker.The results were as follows:1) For all test signal frequencies, the masked threshold increased with masker bandwidth at a rate of 3dB/oct and the upward spread were obtained for supracritical masker bandwidths in simultaneous masking.2) The data for forward masking were different from the data for simultaneous masking in that the slope for subcritical bandwidths were less steep and the data for backward masking were much flatter. Despite this difference, a transition in slope could be identified at approximately the same band width for two types of nonsimultaneous masking.3) The masker bandwidth in that the same suppression effect was obtained as a function of test signal frequency for nonsimultaneous masking were liner and proportional to critical band. These data suggested that suppression effect was possibly due to interaction of auditory filter output.4) For backward masking, the most striking features of suppression effects were the large decrease in threshold for supracritical masker bandwidths and for intermediate intensity, and the greater suppression area of the lower frequency band than higner frequency band. These data showed also that there was much greater suppression for lower frequency band in backward masking than in forward masking.5) These differences were consistent with hypothesis that suppression effect in forward masking was primarily determined by cochlear processes and that suppression effect in backward masking was dominated by central processes, possibly dorsal cochlear nucleus.