Subject: Re: TL-8 Holo tanks?
[email removed] (Richard Johnson) writes:
>The August 13, 1990 Electronic-Engineering Times has a neat article
>about TL-8 Holo tanks. Two prototype systems were displayed at
>Siggraph in Dallas (TX).
...
>So there it is: the first holotank. A monochrome image and high-powered
>computer hard and software. Just think of the MT simulation we could
>write for THAT interface. :-)
Take a look in Newman & Sproull, Principles of interactive computer
graphics, 1973 (1979 2nEd) (McGraw-Hill): there have been repeated
attempts to make the rotating mirror, the rotating plane of LEDs,
and similar things work. Another early attempt at MITRE involved a
flexible mirror stretched over a speaker running 60hz tone.
Bolt Beranek & Newman has recently exhibited an improved model of
the MITRE moving mirror 3d scope which is now for sale at reasonable
prices (was it $10k in quantity 10?) for monochrome wire-frame or
points. They got around MITRE's problems by using custom hardware
to compensate for the varying focal length of the mirror and
resulting paralax of points painted rapidly on the CRT and to do the
(x,y,z) -> (t-scan,h-scan,v-scan) coordinate transform. Is was at
the demo a year or so ago, sponsored by siggraph/ne. I believe the
FAA bought a few for experiments in air traffic control.
If you buy a sheet of mylar & strech it over an old 10' woofer and
gaze 'through' it into fast-phosphor monochrome tube, you can do
this as a hobbyist -- the uncorrected transforms are bearable, they
say.
>Of course hard-copy is another problem...
Also currently in production. I forget the company's name, but they
call their 3d output devices, marketed for CAD/CAM services,
'StereoLithograhpy', which I find misleading. But is literally is a
tank. Their technique is to fill a tank with a polymer liquid which
when given sufficient energy boost will form a solid polymer. Two
tuned lasers scan a surface in the tank; if both are ON, the light
energies at their intersection is sufficient to cause a local
reaction converting liquid to solid. The bottom of the tank slowly
drops, moving a new plane into the construction zone.
It is limited to constructed objects which can be self-supporting
during construction. Current software results in a fairly grainy
surface I'm told, subject to jaggies. This will surely improve.
Typical users take the produced prototype and either cast a mold
from it (master negative) or use lost-wax process to cast a master
positive from which to make molds, after polishing the master.
I think I saw a net article on this, I'll see if I can find it
& forward it along.
If you don't insist on additive technology for 3D 'tank' hard-copy,
there are a number of CAD-driven robotic routers/sculpters
chipping/grinding away everything that doesn't look like the CAD
drawing. :-)