Shrinkage of perceived tonal duration produced by extra sounds: effects of spectral density, temporal position, and transition direction

To investigate the mental mechanism that estimates the duration of sounds, the subjective duration of a tone was measured. In the first experiment a portion of the target sound was replaced with another sound. In the second and third experiments another sound either started or ended in the middle of the target sound. Ten or eleven undergraduates participated as listeners in each of the experiments. In the first experiment, effects of spectral spacing and those of temporal position of the replacing sound were tested. Compared with the intact case, the subjective duration of the replaced target shrunk, and the degree of shrinkage increased as the spectral spacing became wider. The temporal position of the replacing sound did not affect the degree of shrinkage. In the second experiment, effects produced by the start (ON) of the concurrent sound and those by the end (OFF) were compared. The ON case was more effective than the OFF case. In the third experiment, effects of the rise time and fall time of the concurrent sound were tested. A long rise and fall time reduced the difference between the ON and OFF cases. These results are discussed from two viewpoints, one assuming an interaction between the time markers, the other assuming continuous gate control in a neural-counting model for duration.