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

Re: Rationale for Critical Bands

Dear Peter, and list,

it looks we can learn a lot from your two references. As to your question if we could adapt to something like 1 critical bands (CB) per octave, I would say no. The number of around 4 CBs per octave in the most important hearing range that humans and cats have, and probably also rats and mice, appears to be a kind of optimum adaptation. More CBs would not add significantly to spectral integration, such as pitch synthesis. Less CBs would weaken the basis for spectral integration.

As to neural plasticity, it is more probable to occur within a neural lamina, and between adjacent laminae, than across several laminae.

The brain has a great capability to invest into improved spectral filtering, as we see in musicians. But there are also anatomical limits. To shift the number of CBs per octave by training does not seem realistic at all.


Martin Braun
Neuroscience of Music
S-671 95 Klässbol
email: nombraun@xxxxxxxxx
web site: http://www.neuroscience-of-music.se/index.htm

----- Original Message ----- From: "Peter Meijer" <feedback@xxxxxxxxxxxxxxxxxxx>
To: <AUDITORY@xxxxxxxxxxxxxxx>
Sent: Tuesday, June 15, 2010 8:58 AM
Subject: Re: Rationale for Critical Bands

Thank you for these very interesting references, Martin.

To the extent that critical bands find their origin in the ICC, one could
conjecture that critical bands may be subject to (some) neural plasticity.
Are there any studies with, say, young animals to see if the distribution
of critical bands can be modified through suitable exposure and training
with sounds, e.g., aimed at narrowing the critical bands? Natural settings
for critical bands would likely already be "optimal" in some sense for
natural sounds, but it would be interesting to know if adaptation is possible
in (artificial) situations where an optimal setting would be different.
Perhaps even a dynamic modulation of critical bands depending on momentary
needs would be possible in cases where that serves the optimality condition.

A quick web search turns up references like


The second of these two references does seem to get close to addressing my
question, but I'm not sure if it is the most representative example. I am
personally most interested in possible effects of plasticity with complex
sounds that are non-harmonic, with close to 20 tones - or more - per octave in a log distribution. Standard critical bands cause a within-band temporal
"roughness" that might play a role in guiding effects of plasticity for a
better spectral envelope discrimination with complex sounds, but I am not
aware of much literature on this so would appreciate to learn more about it.

Thanks and best regards,

Peter Meijer

Seeing with Sound - The vOICe

Date:    Mon, 14 Jun 2010 15:12:16 +0200
From:    Martin Braun<nombraun@xxxxxxxxx>
Subject: Re: Rationale for Critical Bands

Dear Dan, and list,

You are quite right. The level independence of the critical bandwidths (C=
over wide sound level ranges excludes an origin in the cochlea.

There are compelling data indicating that the origin is in the central=20
nucleus of the inferior colliculus (ICC).