[sdiy] The eardrum and brownian motion

[Chris Stecker]

> Here's a notion I have had for a while. The ear can
> tell the shape of the room just by the ambient noise.
> You can tell is you're in a cave just by the sound.
> Perhaps this is a leftover survival trait? Kinda like
> bat echo-location? If so, then ambient noise levels
> would affect the ear's ability to do this, and lack of
> noise would disable it completly.

I've never had anyone report such a negative experience in the chamber.  It 
sounds more physiological than psychological.  Are you certain that there is 
adequately airflow to insure oxygen availability in the room?  No gas leaks, 
etc.?  

OTOH, I'd imagine such effects could indeed come from disorientation in the 
room.  I wouldn't expect the "lack of sound" idea to hold much water, though; 
working in the room, I'm sure you generate a good deal of sound, unless 
you're very careful to suppress breathing noise and work very quietly 
(tiptoes, etc.)   Unlike some researchers in the area, I _do_ believe that 
people use acoustics to determine some aspectes of the room geometry.  This 
could involve self-generated sound as well as ambient sound. However, I 
think the problem in an anechoic chamber is that there are no reflections, 
not really the low ambient noise level per se.  Without reflections, there 
are no spatial cues about the room, so if there is a mechanism that's trying 
to determine the room shape, it's foiled.  Additionally, an important cue for 
the perception of distance is the ratio of direct to reflected sound.  In the 
chamber, this ratio is abnormally high, so sounds may appear to be MUCH 
closer than they actually are.  Some disorientation might be expected when 
something (another person, for example) sounds very close (like whispering in 
your ear) but is seen visually as some distance away.

One more thing to consider:  I'm sure many of you know about the precedence 
effect, demonstrated classically by presenting two identical sounds from 
different location and with a short delay (1-10 ms, but depends on the type 
of stimulus) between them.  Listeners hear the pair as a single sound located 
in the direction of the early copy.  Researchers often take this effect as 
evidence of "echo suppression," because the delayed copy is basically a 
virtual echo and is perceptually suppressed.  Well, the presence of an 
independent noise source in the chamber reduces the precedence effect, so 
that the late copy is perceived correctly as a second source. You might think 
of the noise as providing information about the room to the listener, proving 
the room to be anechoic.  Since the room is anechoic, the paired stimulus 
cannot be a source and echo, and is instead heard as two separate sounds.  
This seems quite supportive of the notion that ambient (or generated) sound 
helps create an "image" of the room.  It might explain why bringing the fan 
in with you helps, by disambiguating the acoustic cues so you "know" it's 
anechoic.

-Chris