Comodulation masking release (CMR): effects of gating as a function of number of flanking bands and masker bandwidth

Normal-hearing subjects participated in two CMR experiments. For experiment 1, two, three, five or nine 20-Hz-wide comodulated flanking bands were presented continuously or gated simultaneously with a 2000-Hz signal. The signal had a duration of 400 ms. Larger CMRs were obtained as the number of flanking bands increased for both the continuous and gated conditions. For fewer number of bands, the average CMR for continuous noise was substantially larger than for gated noise. As the number of bands increased, CMR increased more for gated than for continuous noise, and the difference between CMRs for gated and continuous noise decreased. Experiment 2 involved detecting a 400-ms 1000-Hz pure-tone signal in a seven-band comodulated noise complex. The noise bands, presented continuously or gated with the signal, were either 10, 20, 40, or 80 Hz wide. Larger CMRs were observed for continuous maskers and for smaller masker bandwidths; however, the effect of gating did not change significantly across bandwidth. The results of the first experiment indicate that the effects of gating on CMR are minimized when the number of auditory channels providing information is large. The results of the second experiment indicate that this effect is not simply a function of the large CMR magnitude obtained with a large number of flanking bands.