Abstract:
One of three pure-tone frequencies was presented in masking noise on a given trial, and the listener was asked to identify which pitch (low, middle, or high) was presented. The level of the signal varied adaptively. In a masking-level difference (MLD) paradigm, the noise was always interaurally in phase (No) and the signal was either interaurally in phase (So) or was interaurally inverted (S(pi)). In a comodulation masking release (CMR) paradigm, a monaurally presented masking noise was either unmodulated or square-wave modulated at a rate of 30 Hz. In each paradigm, the parameter was the frequency separation between the signals. Detection thresholds were also measured for each frequency. When the frequency separation was relatively wide (e.g., 100 Hz or more), detection thresholds were similar to identification thresholds. As the frequency separation decreased, the identification thresholds increased steeply, particularly in the masking release conditions (NoS(pi) and modulated noise). Results will be discussed in terms of the different cues available in the baseline and masking release conditions, and in terms of why those cues may be less informative in identification than in detection conditions. [Work supported by NIDCD.]