Multiple timescales of context influence perceptual sensitivity to common pairings of musical pitch and timbre.

Previous studies have established that musical pitch and timbre (specifically, spectral shape) perceptually covary: lower pitches are associated with darker timbres (less higher-frequency energy) and higher pitches are associated with brighter timbres (more higher-frequency energy). In four experiments, perceptual sensitivity to this relationship was assessed in pitch labeling tasks when instrument timbre varied in ways that respected or violated this pattern (Consistent or Reversed trials). Performance was influenced by context at multiple timescales: block-level (stimulus type), experimental session-level (block order or configuration), and longer-term experience (musical training background). Across experiments, participants performed near ceiling accuracy for Consistent stimuli, but were less accurate for Reversed stimuli. This pattern was moderated by which condition was tested first in the experiment, the introduction of trial-by-trial feedback, and presentation of trials in blocked versus interleaved orders. Higher musical training scores were generally associated with higher accuracy on Consistent trials but were more reliably and more strongly associated with higher accuracy on Reversed trials. Thus, context on multiple timescales can shape perceptual sensitivity to the natural covariance between musical pitch and timbre. Results advance the efficient coding hypothesis by demonstrating how listener factors can modulate perceptual sensitivity to statistical structure in the acoustic environment.