Growth of forward masking for sinusoidal and noise maskers as a function of signal delay; implications for suppression in noise.

The first two experiments were designed to determine whether mutual suppression in broadband noise increases in strength with increasing overall level. In experiment I masking functions (signal threshold versus masker level) were measured in forward masking as a function of the delay time of a 10-ms signal, both for a broadband noise masker (low-pass filtered at 8 kHz) and for sinusoidal maskers at 1, 2 and 4 kHz. In the latter case the signal frequency equaled the masker frequency. For short signal delays the masking functions were steeper for the sinusoidal masker than for the noise masker. At longer delays the slopes for both masker types decreased and the slopes for the two masker types became more nearly equal. In experiment II we investigated the effect of gating a low-level noise cue with the sinusoidal masker. At the longer signal delays the masking functions had equal slopes for the broadband noise masker and the sinusoidal masker with cue. At short signal delays the masking functions for sinusoidal maskers may be "artificially" steepened, since the subject lacks an effective cue to distinguish the signal from the masker. The equal slopes at longer delays indicate that mutual suppression of the components within a broadband noise does not increase in strength with increasing overall level. In experiment III we attempted to estimate the magnitude of mutual suppression in a broadband noise by comparing masking functions for a broadband noise and for a noise whose bandwidth was 20% of the center frequency. The suppression was estimated to be about 2 dB at 4 kHz and 8 dB at 2 kHz. A simple mathematical expression, suggested by Jesteadt et al. [J. Acoust. Soc. Am. 71, 950-962 (1982)], was found to give an accurate description of the amount of masking produced by the broadband masker as a function of masker level and signal delay.