Detection and identification of a single modulated carrier in a complex sound

Hall and Grose [J. Acoust. Soc. Am. 90, 3028-3035 (1991)] reported that subjects had difficulty in deciding which carrier in a two-carrier complex sound was modulated. The present experiments examined how the ability to identify a single modulated carrier was affected by the number of carriers in the complex and by harmonicity. Initially, thresholds were measured for detecting amplitude modulation of a single carrier in a complex sound. Thresholds were higher when that carrier was one of the inner carriers in a six-carrier harmonic or inharmonic complex than when it formed part of a two-carrier complex. Thresholds were only slightly, if at all, higher when the modulated carrier was varied randomly from trial to trial than when its frequency was fixed within a block of trials. Next, subjects were required to decide whether the frequency of a single modulated carrier (with a suprathreshold modulation depth) in a complex sound was the same as or different from the frequency of a probe composed of a single modulated carrier. They generally performed well above chance. Performance was not greatly affected by whether the probe was presented before or after the complex, but was generally slightly better for a modulation depth of 100% than for a depth of 50%. Randomly varying the level of each carrier in the complexes from one stimulus to the next produced only a slight impairment of performance, indicating that short-term across-frequency differences in level were not used to identify the modulated carrier in experiment 2. Overall, performance was best for the six-carrier harmonic complex, less good for the six-carrier inharmonic complex, and worst for the two-carrier complex. The results are interpreted in terms of perceptual grouping.