Evidence for a general template in central optimal processing for pitch of complex tones

Area of Science:

• Auditory perception
• Psychoacoustics
• Signal processing

Background:

• The optimum processor theory explains pitch perception by assuming component tones are successive harmonics.
• This leads to a paradoxical prediction: complex tones with non-successive harmonics may not evoke their fundamental pitch.

Purpose of the Study:

• To test the validity of the successive harmonic constraint in periodicity pitch perception.
• To develop a more general model for estimating periodicity pitch.

Main Methods:

• Conducted new pitch-shift and musical interval recognition experiments.
• Re-analyzed existing pitch perception data.
• Modified the optimum processor theory by replacing maximum likelihood estimation with maximum posterior probability estimation and removing the successive harmonic constraint.

Main Results:

• Experimental results demonstrate that the successive harmonic constraint is not universally true.
• The revised theory successfully reconciles new and old experimental data.
• Periodicity pitch is estimated by matching stimulus components to a general harmonic template.

Conclusions:

• The successive harmonic constraint is not a general requirement for periodicity pitch perception.
• A more generalized optimum processor theory, using maximum posterior probability estimation, provides a better account of pitch perception data.
• This revised model encompasses the successive harmonic constraint as a specific instance.