A unitary model of pitch perception

Area of Science:

• Auditory Neuroscience
• Psychoacoustics
• Computational Auditory Processing

Background:

• Theories of pitch perception have historically diverged, particularly regarding resolved versus unresolved harmonics.
• Understanding the mechanism of residue pitch perception is crucial for explaining complex auditory processing.

Purpose of the Study:

• To revisit and validate a computational model for residue pitch perception.
• To demonstrate the model's ability to reconcile historical theories of pitch perception.
• To explain how harmonic resolvability influences pitch percept quality.

Main Methods:

• A four-stage sequential processing model was employed: peripheral frequency selectivity, within-channel processing (rectification, filtering), within-channel periodicity extraction (autocorrelation), and cross-channel aggregation.
• The summary autocorrelation function (SACF) was used to represent pitch information.
• The model's simulation capabilities were tested against new experimental results.

Main Results:

• The model successfully simulated new experimental data on pitch perception influenced by harmonic resolvability.
• Specific phenomena addressed include phase alternation effects, increased frequency difference limen for higher harmonics, and mistuned harmonic influence.
• Comparing stimuli along the SACF, not just the highest peak, was essential for model accuracy.

Conclusions:

• The revisited model can reconcile various historical approaches to pitch perception theories.
• The model demonstrates differential responses to low and high harmonic complexes.
• A single mechanism can explain pitch percepts for both resolved and unresolved harmonics, eliminating the need for dual-mechanism theories.