Forward masking with equal-energy maskers

Abstract

Temporal masking of chicks by noise was investigated using a forward-masking paradigm. The temporal separation delta T between the click and the noise ranged from 0.03 to 100 msec. The duration of the noise varied from 3 to 500 msec while its energy remained fixed. For fixed delta T (delta T greater than 3 msec), the masking effect may actually increase for the longer, less intense noises despite the fact that, for long maskers, there is less masker energy near the signal in time. These results are interpreted in terms of the modified version of the running-average hypothesis [M. J. Penner, J. Acoust. Soc. Am. 63, 195--201 (1978)] in which it is argued that the auditory system compresses the intensity of the stimulus prior to integrating it. If the temporal integrator depends on stimulus intensity, then these results may be easily predicted. As an alternative explanation we show that compression may reduce the effective intensity of short maskers to such an extent that they do less masking than the longer ones. Such reduction in masking effectiveness will occur if the time between the masker and the signal is long enough so that the effects of compression and integrator shape do not counterbalance each other.