Spectral fusion based on coherence of amplitude modulation
Stanley Sheft, William A. Yost · The Journal of the Acoustical Society of America · 1989
The ability of listeners to attend to a subset of components (the target) of a tonal complex was investigated using a forced-choice procedure. The stimulus was an eight-component tonal complex. The components of the target subset (n = 1, 2, or 3 components) were distinguished from the complex by coherent amplitude modulation (AM). Each trial of the cued 2IFC task was preceded by a presentation of the coherently modulated target components in isolation. For each interval of a trial, either the target or an equal number of nontarget components shared the coherent AM. Subjects were required to detect the interval in which the target components were coherently modulated. The remaining components of the complex were all either modulated at different rates, modulated with a random shift of the target-modulator phase angle, or not modulated. Performance was measured as a function of the number of target components, the harmonic relationship among the target components, and the harmonic relationship between the target and nontarget spectral groups. With coherent AM of harmonic target components, the increment in d′ with increasing n exceeded predictions based on the combination of independent sources of information, suggesting that these stimuli may be processed as an entity by the auditory system. [Work supported by NINCDS.]