Perceptual segregation of a harmonic from a vowel by interaural time difference and frequency proximity.

The five experiments reported here examine the conditions under which sounds differing in their interaural time difference (ITD) are segregated for the purposes of perceiving a vowel's identity. Experiment 1 confirms previous findings that (i) a difference in ITD provides only a very weak cue for segregating a vowel's 500-Hz harmonic from the remainder of an isolated vowel; (ii) embedding the harmonic in a series of 500-Hz tones produces some segregation, which is enhanced if the harmonic and the tones differ in ITD from the rest of the vowel; and (iii) when these latter sounds are presented in the same block as isolated vowels, they facilitate segregation of the harmonic by ITD in the isolated vowels. The subsequent experiments show that this last effect, across-trial facilitation, is only produced by sounds which cue both the frequency and the ITD of the harmonic; either alone is insufficient. We also show that: (i) a single cue tone at the frequency of the harmonic is sufficient to facilitate the use of ITD in grouping; (ii) sequential organization by frequency proximity dominates over sequential organization by ITD when simultaneous sound sources are present; and (iii) the effectiveness of a cue tone can be abolished by capturing it into a synchronous harmonic complex. The experiments clarify the conditions under which ITDs contribute to the segregation of simultaneous sounds.