Effects of the selective disruption of within- and across-channel cues to comodulation masking release

Abstract

In many experiments on comodulation masking release (CMR), both across- and within-channel cues may be available. This makes it difficult to determine the mechanisms underlying CMR. The present study compared CMR in a flanking-band (FB) paradigm for a situation in which only across-channel cues were likely to be available [FBs placed distally from the on-frequency band (OFB)] and a situation where both across- and within-channel cues might have been available (proximally spaced FBs, for which larger CMRs have previously been observed). The use of across-channel cues was selectively disrupted using a manipulation of auditory grouping factors, following Dau et al. [J. Acoust. Soc. Am. 125, 2182-2188(2009)] and the use of within-channel cues was selectively disrupted using a manipulation called "OFB reversal," following Goldman et al. [J. Acoust. Soc. Am. 129, 3181-3193 (2011)]. The auditory grouping manipulation eliminated CMR for the distal-FB configuration and reduced CMR for the proximal-FB configuration. This may indicate that across-channel cues are available for proximal FB placement. CMR for the proximal-FB configuration persisted when both manipulations were used together, which suggests that OFB reversal does not entirely eliminate within-channel cues.

Modeling within- and across-channel processes in comodulation masking release

“…Such effects have been observed in CMR experiments using dichotic noise (e.g., Hall et al, 1990;Goldman et al, 2011) as well as in pitch perception experiments (Bernstein and Oxenham, 2003;Houtsma and Smurzynski, 1990). In the present modeling study, however, the assumption of a frequency-independent process has led to a reasonable agreement between data and simulation.…”

Comodulation masking release: Effects of training and experimental design on use of within- and across-channel cues

“…This effect was not apparent for the third session perhaps because by then the subjects had learned to use the cues effectively. Grose et al (2009) and Goldman et al (2011b) have also argued that auditory grouping processes can affect the ability to use within-channel cues, although in their studies with trained subjects, the specific manipulations used led to poorer rather than better use of within-channel cues.…”

“…Performance was compared between the two halves of the experiment to determine if any training effects had occurred, given the substantial inter-and intra-subject variation across individual sessions (Experiment One) and as observed previously (e.g., McFadden and Wright, 1992). One of the auditory grouping manipulations used in Experiment One (postcursors) was also used here, to allow observation of any differential effects of training on the use of within-versus across-channel cues (for a discussion of the relative merits of postcursors as an auditory grouping manipulation, see Dau et al, 2009;Goldman et al, 2011b).…”

“…The magnitude of CMR was specified as the threshold obtained using uncorrelated FBs minus the threshold obtained with comodulated FBs [CMR(U-C)] (Schooneveldt and Moore, 1987). Discussion of different measures of CMR can be found in Verhey et al (1999) and Goldman et al (2011b). In a "broad" configuration, the FBs were centered at 250, 500, 2000, and 4000 Hz, frequencies chosen to minimize the likelihood that within-channel cues could be used (Piechowiak et al, 2007).…”

“…Differences in training or prior experience across subjects may partly account for the substantial intra-and intersubject variability that has sometimes been reported for CMR experiments (McFadden, 1987;McFadden and Wright, 1992;Grose et al, 2009). Both the reliability of experiments on CMR and the broader interpretation of the results depend upon the degree to which the use of withinand across-channel cues is affected by the amount of training subjects receive and the experimental design employed (Fitzgerald and Wright, 2011;Goldman et al, 2011b). For example, in many studies of CMR, the experimenters themselves were participants.…”