Effect of Bandwidth of Masking Noise on Detection of Homophasic and Antiphasic Tonal Signals

Abstract

Langford and Jeffress [J. Acoust. Soc. Am. 36, 1455–1458 (1964)] report evidence that the bandwidth involved in masking an antiphasic signal, N0 Sπ, is considerably wider than that involved in monaural or homophasic masking. The present paper describes a study of the effect of narrowing the noise band used for masking a 500-cps tonal signal from 100–3000 cps to about 10 cps in width. The same band of noise was employed for both the diotic conditions N0 S0 and the antiphasic N0 Sπ. In the former case, no improvement in detection occurred until the bandwidth was less than about 150 cps, but in the latter substantially increased MLD's were found for bandwidths as great as 300 cps. The rate of improvement of detection under the antiphasic condition was much more rapid as the bandwidth was narrowed than was the case for the diotic condition. The findings support von Békésy's idea that some kind of neural “funneling” assists in narrowing the band of frequencies involved in monaural detection. In antiphasic detection, the bandwidth may depend solely on the mechanical filtering provided by the cochlea without assistance from neural funneling. [Work supported under a grant from the National Aeronautics and Space Administration.]