Chapter 5 of Moore (1997)

現在位置: ROM首頁 / 聽覺心理學 / Chapter 5

{content}

Chapter 5 of Moore (1997)
Pitch Perception(II)

Scope

Definition
Two Types of Theories of Pitch Perception
Frequency Discrimination of Pure Tones: What and How to Measure
Pure Tone Data and Implications
Other Pure Tone Perception Phenomena
The Pitch Perception of Complex Tones: Some Demos and Reviews on the Basic Concepts for Our Discussion Next Time

Definition

"That attribute of auditory sensation in terms of which sounds may be ordered on a musical scale" (American Standards Association, 1960).
A pitch value is generally specified by the frequency of a pure tone having the same subjective pitch as the sound.

Theories of Pitch Perception

Place theory: (1) Following Ohm's acoustics law.
(2) Position of maximum excitation.
Generally speaking, at high frequency range, place of excitation is the
only available code for frequency.
Temporal theory: The time pattern of the neural impulses evoked by the
stimulus. Phase locking in neural response is reliable up
to about 5kHz.

What and How to Measure in Studying Frequency
Discrimination

Temporal 2AFC: -> difference limen for frequency (DLF) (Please run the demo experiment to collect your own DLF's for various frequencies.).
Amount of FM modulation: FMDL.

Data: Pure Tone Discrimination (I)

Fig. 5.1 & 5.2
Zwicker's model: Change in excitation pattern on the low frequency side. See Fig. 5.3.
Henning's control for the loudness cues for frequency discrimination: Random level assignment. -> Only agrees with Zwicker's model at high freq's.
So far, a pure linear model seems adequate to account for the masking data, but….

Data: Pure Tone Discrimination (II)

Moore's method: Stimuli with a cretain spectral slope. The shorter the pulse, the shallower the slope. See Fig. 5.5.
Another alternitive: Moore & Sek's "multi-channel excitation pattern model" -> could account for MM at about 10 Hz rate, but not those at lower rate.
Overall, DLF & FMDL for very low modulation rates are determined by phase locking. FMDL (and to a lesser extend, DLF) at higher rates can be accounted for by changes in the excitation pattern.

Other Phenomena

Tone chroma
When f > 2.5 kHz, octave match becomes difficult.
A sequence of pure tones above 5 kHz do not give good sense of melody.
Even people with absolute pitch cannot name notes above 4-5 kHz.
The pitch of tones < 2 kHz decreases with increasing sound level, while that of tones > 4 kHz increases (What color phenomenon can be viewed as the visual counterpart?).
Does the level-dependent pitch shift favor one competing theory over another?

Here Comes My Favorite: The Case of Missing Fundamental

Helmholtz's account.
What's distortion product?
Early attempts to test Helmholtz's view.
How to generate a missing fundamental complex tone without digital computers?
For me, the development of ideas and implementations along this line of research is one of the most beautiful stories in the study of perception.

If you need a print friendly verison of this document, you can download the .DOC file.