From: John Larkin
Subject: Re: Coming soon to a theatre near you - endless mischief potential
Date: Sun, 13 Oct 2002 20:39:07 -0700
Organization: Posted via Supernews, http://www.supernews.com
X-Newsreader: Forte Agent 1.91/32.564
On 12 Oct 2002 13:37:03 -0700, Winfield Hill
>> even ignoring the Popular Science hyperbole ...
> I'm telling you the technique is perfectly viable and even
> practical, especially at frequencies above 100kHz and longer
> distances. That said, you'll not rope me into defending the
> Popular Science version or their claims. :-) But this
> "field" does seem to draw charlatans out of the woodwork.
> - Win
OK Win, how does this sound:
The inherent pressure-volume behavior of a gas allows nonlinear
effects, specifically the mixing of two ultrasonic waves to produce an
audible difference wave. I assume (because you said so) that the
initial sonic wave is launched in the direction of the parent waves,
given two collinear waves from a single emitter.
Once created, the sonic wave is not constrained by the ultrasonic
beams, so propagates according to the usual rules of wave behavior.
The conversion process is nonlinear (or else it wouldn't convert!), so
conversion efficiency is high nearest the source and drops rapidly
with distance, as the beams spread geometrically. The ultrasonics are
also attenuated by atmospheric absorption, increasing the tendency for
conversion to drop with distance. At 30-50 KHz (which both parties
seem to like) the geometric effect probably dominates. So most of the
sonics are created close-in.
Conversion efficiency is low at practical power levels. Assuming
(waves hands here) 100% conversion efficiency at 1 bar (conservation
of energy upper limit) and a square-law-on-SPL behavior (waves feet,
too) conversion efficiency at 140 dB per beam would be (very) roughly
1 PPM (can that be right? maybe not, but still damned low.)
Since most conversion happens near the source, low frequencies will
begin diffracting there, where they are too big for the generator
size, while the highs will tend to stay more with the beam as it
expands. Diffraction is messy here, since the parent beam does not
have a flat energy profile (it's gaussian or something, or at least
falls off on the edges, and has bumps and lumps and sidelobes in real
life) and, as noted, conversion is nonlinear on local intensity.
Sounds defocussing to me...
There is some harmonic distortion associated with the conversion. A
major factor here would be the inhomgenity of ultrasonic intensity
within the active conversion volume. Since at least one of the
ultrasonic sources is modulated (SSB or whatever) the intensity
patterns are dancing in space in real time (complicated by the
transducer behavior) so I won't go into this for free, and probably
not if you paid me. Some precompensation is apparently possible.
A real-life foot-or-so-diameter planar source will indeed have a hefty
spread angle and bumps and lobes and stuff. Holosonics cites 3-5
degree beam widths, and AmEngr is totally vague on this. Holosonics
proposes 'whispering into the ear of a single person in an audience'
and AE talks of a lifeguard's bullhorn that addresses a single person
on a crowded beach. Not likely. The laser pointer is maybe a bit of
[A really good, large, phased-array source might create a beam that
tapers down to a narrow waist, where intensity would be highest. That
would be interesting. But the narrow waist would still blow the lower
frequency stuff in all directions. Other beam-profile tricks could be
So the thing looks like, in effect, a short invisible horn in front of
a regular loudspeaker, sort of like the things cheerleaders use, with
lots of attenuation and some added harmonic distortion. Cute, maybe a
good vehicle for raising money, but not capable the outlandish claims
made by either company.
One would have to be careful evaluating this, as a very convincing
demonstration could be easily faked, especially for a single observer.
I don't know how the Elephant Radios fit in here.
hey, this was actually fun... haven't done any acoustics in years!