Abstract:

The detection advantage associated with masker envelope coherence across frequency has typically been described in terms of comparisons across auditory channels. More recently it has been suggested that analysis of the output of a wider initial filter, similar to that suggested for the TMTF, can account for the data [Berg, J. Acoust. Soc. Am. 100, 1013--1023 (1996)]. In particular, a change in envelope beats was proposed as the cue to the addition of a pure-tone signal. Data are presented for the detection of a tone added to multiple maskers with coherent envelopes. In one condition a change in envelope beats is an accurate potential cue, and in others it is a much less reliable or unreliable indicator of the presence of the signal. All conditions employing maskers with coherent envelopes produce very similar thresholds, and all showing improved sensitivity over the case of detecting a signal added to a single masker centered on the signal frequency. Results are interpreted as evidence that a change in envelope beats does not form the basis of detection in CMR. One version of the cued-listening hypothesis, an across-channel model, is reconsidered in light of these results. [Work supported by NIH-NIDCD.]