[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Wavelet discussion

"Richard F. Lyon" <DickLyon@ACM.ORG> wrote:

> The main reasons to NOT use a wavelet formulation for the basilar
> membrane are:
> 1) wavelets are inherently constant-Q, but the BM is between
> constant-Q (near the base, or high frequencies) and constant-BW
> (near the apex, or low frequencies).
> 2) wavelets are inherently linear, but BM filtering is strongly
> nonlinear--compressive and adaptive.
> Jont also points out that wavelet users largely ignore frequency
> response.  This is true, but non inherent.

A fundamental question in this context is: "Do we consider a
given property of the human auditory system a bug or a feature?"

Of course, for some purposes this question is simply irrelevant,
such as for cochlea implants where physiological congruence is
essential. However, when it comes to acoustic signal separation
one might ask if a certain bug/feature is worth the effort of
being implemented and what exactly would be gained from doing so.
This might be why some researchers tend to ignore details like the
BM filter asymmetry, just as long as nobody tells them why they
must not be neglected.




 -----------  Ludger Solbach, Distributed Systems Department  -----------
|           Technical University of Hamburg-Harburg, Germany             |
|       e-mail: Solbach@TU-Harburg.d400.De, Tel.: +49-40-7718-3357       |
 ----------------- http://www.ti6.tu-harburg.de/~ti6ls/ -----------------