Re: Lower frequency limit for pitch perception

[Richard M Warren <rmwarren@xxxxxxxxxxx>]

On Wed, 26 Apr 2000, Paul von Hippel -- Ohio State wrote:

> Below a certain frequency, periodic sounds are no longer perceived as having pitch; instead, they have a rough or pulsating quality (these are probably not the right words). But where exactly is this lower boundary for pitch perception, and on what signal qualities (besides frequency) does it depend? I'm guessing the threshold for simple tones is about 25 Hz, but maybe it's different for complex tones? In any case, it would be nice to know what the research literature has to say.
>
> I've looked in a few standard reference books (Moore, Yost & Nielsen, Handel), and searched PsychINFO, but have had no luck -- possibly because I don't know the proper keyword.
>
> Thanks for any help you can provide,
> Paul von Hippel
> Post-Doctoral Fellow
> School of Music
> Ohio State University
> 1866 College Road
> Columbus, OH 43210
> 614 292-7321
>
> von-hippel.1@ohio-state.edu
> (You can also reach me by simply replying to this message.
> All my mail is forwarded to the same final destination.)
>
> http://music-cog.ohio-state.edu/~pvh
>


You are quite right about the approximate boundary for hearing pitch with

simple sinusoidal tones.  Below roughly 20 Hz, pitch cannot be heard.

This is not surprising since the sinusoidal tone becomes inaudible.

However, you ask if it is different for complex tones.  It is -- any

complex tone with harmonic (i.e. sinusoidal) components within the audible

range can be heard when the fundamental frequency is below 20 Hz, but the

percept, though clearly periodic, is not one of pitch.  Excellent stimuli

for studying periodicity detection in both the pitch and infrapitch

ranges are iterated segments of Gaussian noise.  These repeating frozen

noise segments (RFNs) were introduced as a research tool by Guttman and

Julesz in 1963.  Repetition of RFNs are heard with ease from 20 Hz to

1 Hz; from 20 to 4 Hz the percept has been described as "motorboating,"

and from 4 to 1 Hz as "whooshing."  RFNs are useful as generic periodic

stimuli because they consist of all harmonics of the repetition frequency

with randomly determined relative amplitudes and phases produced by the

particular stochastic waveform.  Since you have been working in OSU's

School of Music, you may find it interesting that in the pitch range,

RFNs from about 100 to 2,000 Hz have clear, especially rich, novel timbres

with the pitch corresponding to the repetition frequency, and the timbre

changing dramatically with different random waveforms having the same

pitch.

     Research within and across perceptual ranges in the RFN

pitch-infrapitch continuum is summarized in my book, Auditory Perception:

A New Analysis and Synthesis" (Cambridge University Press, 1999, see

chapter "Pitch and Infrapitch" pp. 56-94).  A CD prepared by Jim Bashford

accompanies the book, and contains demonstrations of RFNs in the pitch and

infrapitch ranges.