The ability of subjects to make same‐different judgments of target melodies interleaved with sinusoidal “masking” melodies with various partials present or absent in the target was investigated. The melodies were stochastically generated with identical melodic centers, limited pitch extent, and no intervals of greater than three semitones. Clearly, the presence of the first, second, or first and second greatly facilitated the subjects performance. This is probably not due to the timbral distinctiveness of the notes in the target, however, because the presence or absence of the fundamental in the target had virtually no effect. This allows the interpretation that subjects ignore the critical band containing the fundamental and attend only to the upper partials. This is consistent with Dowling's finding that the perception of interleaved melodies is greatly enhanced if their pitch separations are approximately a critical band or more [W.J. Dowling, Cognitive Psych. 5, 322–337 (1973)]. [This work was supported by the National Research Council of Canada.]

This content is only available via PDF.