DPRG List  

DPRG: BOUNCE dprglist@dprg.org: Non-member submission from [Gabe Velez <gvelez@richmond.infi.net>]

Subject: DPRG: BOUNCE dprglist@dprg.org: Non-member submission from [Gabe Velez <gvelez@richmond.infi.net>]
From: Alan Bredon alanb at cryogen.com
Date: Mon Jan 5 10:51:24 CST 1998

Gabe,I'm having a hard time keeping my civility after reading your
scientifically inaccurate and confused comments to my suggesstions about
homing in on sound.  I can't tell, from any of your comments, whether
you know anything or not.
Maybe you're just a high-school kid trying to fit-in.

Thanks to Kevin for his *intellegent* comments to my suggestions.

Firstly, the human ear is way more complex in both structure and
ear/brain interaction than it appears.
Humans can not only percieve the direction of a sound but can also pick
out a speaker from a crowded/noisy environment.  This is because the
brain uses subtle phase information presented as a result of reflections
>from the outer ear as well as from the room and objects in it.  The
outer ear in humans as well as most animals is far more complex than a
simple parabaloid.  The shape, depth, and convolutions are all utilized
to characterize and sometimes correct the hearing response.

Trying to duplicate this complexity would be difficult and would require
DSP and some parallel processing, but it is possible.  I merely made
suggestions in direct response to someone who needed a *simple* way to
home a robot to command sounds.  Any environment in which a robot is
operating is going to be noisy, busy, and far from controlled.
Experience shows that simple is sometimes better.  To try to
microanalyze sound commands would prove to be very difficult in that
noisy environment.  Better results might be acheived by looking at sound
patterns and echo timings.
A directional microphone system is far easier than trying to look as
soundwave phasing.  The noise would screw it up.

That brings up another idea on directional sound detection.  This was
used on the Heathkit Hero, I believe:
Use a "mirror" reflector attached to a stepper (or servo or motor with a
potentiometer) to rotate 360 degrees, reflecting the the sound down to a
mic below (facing up).  I've been planning to use this approach on my
Otto.  This reflector is flat and mirrored to also reflect light and the
poloroid ranging signals.  The Hero used this technique to "follow" a
flashlight, to track sounds, and for ranging.  The reflector's position
(azimuth) is the sound/light position.  It can spin continuously (like a
radar does) or use a controlled rotation algorithm.  I plan on adding an
elevation component to allow ranging to the upper walls for navigation
and lower for obstacle detection.  Ranging, navigation, sound and light
tracking, intrusion detection, etc. all in one unit! Cool or what?

There has been some recent research on our ear's ability to use phase
information to determine sound direction.

Also you need to look into the technology used in the Surround Sound and
3D Sound generating equipment.
These use complex phase information to provide sound that has *depth*
and more direction which can actually place sound objects in specific
places around you, all with 2 or 3 speakers.

To say that a cardioid pattern is not needed is simply an ingnorant
flame-inducing faux pax.

>From the rest of your comments about "analog buffers" and "I think the
motion circuit should be set for the threshold." it is obvious that your
association with the neutronicstechcorp and CORE scammers has completely
fried your brain.

> >Subject: BOUNCE dprglist at dprg.org:    Non-member submission from
> [Gabe
> Velez <gvelez at richmond.infi.net>]
> >    Ah, but notice what you yourself said, "the parabolic shape of
> the
> >ear lobes". The canal and ear mechanism itself is not cardioid.

BTW: Ear Lobes are not parabolic, they are the protrusions of skin on
which earings are hung.

> >    This is only because we have eyes and hair. Tests done using
> >headphones (eliminating acoustic dampening of hair, ears, etc.) or a
> >blindfold (eliminating whether the person can see if  it is actually
> in
> >front of him or not) show that a person cannot detect whether a sound
> of
> >an object or noise is from behind.  But I did not do the tests.
> Acoustic
> >engineers, psychoacousticians, et al, have. But I just read about it.
> I
> >just tried it on myself to prove it to only myself.

- --
     _  _
    / || \    Regards, Alan Bredon
   /  ||  \   My views & opinions do not represent my company.
  /. .||. .\  My email address is MINE.  Spammers should pay me!
 /-|-||||-|-\ Legit replies delete the -NOSPAM.
 | . .  . . | And always remember to ESCHEW OBFUSCATION!


More information about the DPRG mailing list