Some Consequences of the Masking Produced by Combination Bands of Different Types

Abstract

Data reported earlier [D. D. Greenwood, 78th Meeting of the Acoust. Soc. Amer. (Nov. 1969)] demonstrated the existence, and masking effects, of combination bands of the same form as combination tones given by (n+1)FL−nFH, where FL, FH are the frequencies of the primaries. Combination bands could be generated by a pure tone and a band of noise, by two bands of noise, or by only one band of noise. In the latter case, the combination bands (n=1,2,3,⋯) overlap and form the combinational aggregate. A relationship of combination tones, or bands, of this type to critical bandwidth has been shown: when two primary stimuli are separated by one critical bandwidth, the first (n=1) combination tone or band stands at a frequency locus coinciding with the low-frequency foot of the masked audiogram produced by the lower primary. Several considerations lead to the inference that critical bandwidth extends downward from the lower primary tone, and support further the view of critical bandwidth as a resolution, and integration, interval. These experiments also reveal the role of combination bands (and the combinational aggregate) in remote masking and permit the demonstration of difference bands corresponding to the difference tones, (FH−FL) and 2(FH−FL). Other data indicate that combination components can be generated iteratively in situations where three external primaries are used.