Audiogram defects (was Re: Cochlea -- quality of filtering)
tony at howl.demon.co.uk
Mon Sep 25 13:29:02 EST 1995
In article <445g2e$5j9 at geraldo.cc.utexas.edu>
sirianni at uts.cc.utexas.edu "Jeffrey Sirianni" writes:
> No one would argue that hair cells and auditory
> nerve fibers are sharply tuned at low intensity levels, but we know
> that we can excite such cells at a frequency well off characteristic
> frequency when we make the intensity loud enough. But we know that
> even though a cell may respond to a stimulus well of CF, we do not
> necessarily percieve a sound with a pitch at CF due to inhibitory
> mechanisms in neurons from the cochlear nucleus and up.
As I understand it, the critical band theory says we can't distinguish
a *pure* tone of a certain characteristic frequency, from one which has
sidebands which excite cells of adjacent CFs less than they would be
excited by the pure tone alone. Once past the cochlea, the pure tone
and the "impure" tone are the same. They would both sound like what
we are used to calling a pure tone.
Normally if you are exciting cells way off their CFs with a pure tone
sound, you are exciting others right on their CFs, so the pattern of
nerve cell firing fits what we are used to calling a pure tone. So
it's not surprising that we "hear" it as a pure tone.
I think Eric Smith's idea was that this might be different if
some hair cells are dead and we apply the pure tone sound at the CF of
the dead cells. There are a few questions raised here. First do you
mean people with absolute pitch - if not, how do you propose to set up
the experiment? Secondly, what model of pitch perception do you
prefer - firing rate or detection of position on the basilar membrane?
I believe that firing rate is the currently favoured theory at least
for pure tones below 4 to 5 kHz. If this is right, then exactly what
hair cell fires shouldn't affect pitch perception of pure tones much.
All interesting stuff, and it's good that it made me blow the dust off
a few books!
Tony Woolf (tony at howl.demon.co.uk)
More information about the Audiolog