Re: First moment of spectrum

[=?X-UNKNOWN?Q?T=F3th_L=E1szl=F3?= <tothl@xxxxxxxxxxxxxxx>]

On Tue, 13 Jun 2000, Pierre Divenyi wrote:

> I am amazed that no phonetician has so far bothered to weigh in in this
> debate, and I have not seen any psychoacoustic concerns addressed, either.
> For one, there are bi- and multimodal spectra in the acoustic world around
> us. Prey tell, what would the first moment tell about an /i/ or /u/ vowel?

Chistovich did vowel identification experiments in which she tried to
explain certain perceptual phenomena (like the fusing of formants when
closer than 3.5 bark) using spectral first moments. Unfortunately, I
cannot recall her results and conclusion, but can give you the reference:
L.A. Chistovich: Central auditory processing of peripheral vowel spectra,
JASA 77(3), March 1985, pp. 789-805.

> The moment should not
> be regarded more than a mere first-approximation to the auditory
> characteristics of complex spectra.
>
Well, of course. The other paper I mentioned (Albesano et al.) calculates
the first moments of approximately the frequency bands of the first three
formants, and uses these as the estimation of the formants. Of course this
is no more than a gross estimation, but it is at least very easy (=fast)
to compute.

               Laszlo Toth
        Hungarian Academy of Sciences         *  "In our life there's if
  Research Group on Artificial Intelligence   *   In our beliefs there's lie
     e-mail: tothl@inf.u-szeged.hu            *   In our business there's sin
     http://www.inf.u-szeged.hu/~tothl        *   In our bodies there's die"