| From: | "Arno R. Lehmann" <arlehma@***.NET> |
|---|---|
| Subject: | Re: Ultrasound vision |
| Date: | Wed, 3 Feb 1999 20:41:44 +0200 |
>On Wed, 3 Feb 1999, Adam Getchell wrote:
>
>> Moving along to technology that will be present, and should be present in
>> Shadowrun, we have interesting research conducted jointly by Scripp's
>> Oceanographic Institute (aka UC San Diego) and the U.S. Navy on "acoustic
>> daylight".
>
> The "acoustic daylight" stuff is cool, but represents a pretty
>serious computational burden. First off, you're looking at sound over a
>variety of frequencies, some of which attenuate quickly, some of which do
>not. The other problem you have is binocular localization. You need an
>array of sensors to be able to determine directionality of the background
>noise. Tiny microphones are all well and good, but there are problems
>giving them enough directional focus to be able to accurately determine
>direction (especially at distance). Granted, packing a lot of them
>together helps, but you still have problems.
Actually, that's less big a problem. Today's sonar systems work
astonishing similar to what Adam described, and they avoid what you see
as a problem: With an array of sensors, you can find the direction of a
source of sound, simply by distinguishing certain specific sounds
through their waveform. Then you go see which sensor gets that signat
at whaich time, and since you know the relative positions you can
easily calculate the origin of the sound (triangulation you call it, I
guess).
Sure, this does need some computational power, but then, this is SR and
we've got the matrix and all this stuff, so I think we can assume that
that would not be a serious problem.
The point is, you actually don't need directional microphones as long
as you've got a good computer.
<snip>
> Passively? How long is the integration time? It's got to be more
>or less outrageous. In SR terms, you don't want to wait 20 minutes before
>you can distinguish "friend" blobs from "foe" blobs, especially
not when
>you consider that the doppler shift associated with moving targets will
>play hell with your resolution.
Once again, what you see as a problem -- dopler shift -- is also giving
information, in this case about the relative speed of your target.
Still, I do assume that SR's computers are powerful enough to need only
short integration times.
Arno
--
This message serves also as a test if my reply-to address is now
overwritten by listserv -- do not complain if it isn't!
Remember to check the reply-to if you answer this message!
