The effects of timbre on perceptual grouping in a melodic sequence

The current investigation sought to examine the effects of timbre on perceptual grouping in melodic sequences. While past research has shown that timbre shifts influence listeners’ pitch perception on a note-to-note basis (e.g., see Pitt, 2004; Russo & Thompson, 2005, & Creel, Newport, & Aslin, 2004), the current investigation extended this to timbre’s influence on pitch perception in the context of a melodic phrase. In Experiment 1, participants were presented with melodic sequences, made of sawtooth-like waves. Sequences, consisting of 6 tones, were followed by a target tone that had a static, dull, or bright timbre shift through the use of low-pass filters in order to shift the spectral centroid. Target tones were equally presented at ascending and descending interval sizes of a minor 3rd, perfect 4th, and minor 6th. These target tones were paired with timbre conditions, equally, to create timbre shifts that were static, where timbre did not change at all, congruent, where timbre and the pitch of the target moved in the same direction, and incongruent, where timbre and pitch moved in opposite directions. Participants were tasked with rating how well the target tone belonged to the sequence before it. Experiment 2 extended a similar approach to instrumental stimuli. Cello samples were filtered so that the corresponding impact on spectral centroids was similar to the timbre manipulation in Experiment 1. Contrary to hypotheses, participants rated target tones as being less likely to continue the initial melody if any form of timbre shift was present, regardless of interval size. This effect of timbre suggests that it was not subsumed by high-order processes of melody perception. As hypothesized, this effect was negatively related to musical training. Additionally, as expected, interval size influenced ratings regardless of timbre shifts, where larger intervals were less likely to be perceived as belonging to the initial melody. Thus, participants also appear to have used expectations about pitch intervals to make judgments. Finally, the direction of the initial interval within the sequence also influenced target judgments when the target tone constituted a shift in timbre, indicating that participants used directional information to create expectations for the target pitch. Taken together, the findings from the current investigation minimally indicate that, at least under conditions reflecting a single change in instrument source, timbre has the capacity to drastically impact the perception of melodic phrase structure