Re: Redefining [a] and [æ]? or using a better vowel quadrilateral?

In article
<9f3bfb96-5eb9-4060-91cb-866c70d96d5f@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
"ranjit_mathews@xxxxxxxxx" <ranjit_mathews@xxxxxxxxx> wrote:

On Aug 10, 7:31 am, "Peter T. Daniels" <gramma...@xxxxxxxxxxx> wrote:

What makes you think that the brain's response to linear scales is
linear?

In the case of formants, the linearity of scales is arrived at by
measuring distances between formants of vowels in people's speech; the
frequencies are clearly in artithmetic rather than geometric
progression. If babies learn to reproduce them, their brains must
respond linearly.

Like human perception of most physical scales, our perception of
frequency is not linear (note, for example, musical scales, with one
octave corresponding to a doubling of frequency). For frequencies in
speech, especially for formants, the Bark scale is often used. It's
approximately linear below 500 Hz, and approximately logarithmic above
500 Hz. There are multiple different formulas used to calculate Bark,
because the scale itself is not rigorously defined: it was determined
experimentally by testing listeners' responses.

Other non-linear scales used for frequency include mel and ERB.

Nathan

--
Nathan Sanders
Linguistics Program
Williams College
http://wso.williams.edu/~nsanders/
.

Quantcast