Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

New 3-D Technology from Canon (--> NeXT?)

9 views
Skip to first unread message

Barry Merriman

unread,
Apr 26, 1992, 6:51:33 PM4/26/92
to
The following report is captured from sci.virtual-worlds.
It contains a press release about exciting new 3-D rendering
software from Canon, which should have a major impact on
NeXT 3D graphics capabilities (namely cheap hi-quality 3-D animation!).


Read on and salivate!

-Barry Merriman

-----------------------------

From: cybe...@milton.u.washington.edu (Robert Jacobson)
Subject: TECH: "Software Workstation" from Canon UK, press release
Organization: WORLDESIGN, Seattle
Lines: 173

This may be the most important development I was fortunate to see while
in Britain. I won't pretend I know how the software works -- that's
Canon's secret for now, pending the award of a patent -- but the
"Software Workstation" demonstration arranged for me was extremely
impressive. Objects running on various platforms, including off-the-
shelf workstations and PCs, were pulled together into a virtual world
environment maintained by another machine. There was even a virtual
world running in ASCII on a Tektronix dumb terminal. I kid you not.

At the time I saw the display, tools had not yet been built into the
software, with the exception of a sound monitor and analyzer. However,
David Lau-Kee, project director, informs me that a complete virtual
office has now been constructed for the purpose of this first public
showing (detailed below) in France.

Canon is not in the computer business per se, so it will be interest-
ing to see how this new product is developed. Suffice it to say that
it creates some interesting dynamics for manufacturers of large work-
stations and proprietary softwares which do not have the capability of
interoperating with this package. Also, proponents of "garage VR" can
now move out of the garage and into development applications. Tele-
virtuality and telepresence applications also beckon due to the low
bandwidth requirements of the networked world.

I append the Press Release from Canon UK, dated 27 April 1992, and
thank the gentlemen from Canon UK for their openness in telling me
about this technology and sharing it with the sci.virtual-worlds
audience.

--------------------------------------------------------------------

Date: Fri, 17 Apr 92 12:04:10 BST
From: lau...@canon.co.uk (David Lau-Kee)
<< Press Release >>

27th April 1992, Rennes, France.

At the official opening of Canon Information Systems
Research and Development Europa (CISRDE), its sister company
Canon Research Europe (CRE) located in Guildford, UK, today
announced a technological breakthrough in the development of
interactive 3 dimensional computer graphics.

Computer graphics applications allow objects to be
displayed on a computer screen. 3D computer graphics allows 3
dimensional objects (called "models") to be displayed from any
angle and position. For example, a model of an aeroplane could
be displayed from above, or in front; even the view outside
from the pilot's perspective could be displayed. Interactive
3D graphics allows the users to change their viewpoint in near
real-time, thus giving the impression that they are
manipulating a real 3D object. Typically, very expensive,
special purpose computers are used to process and display
interactive 3D graphics, thus limiting its use to high-cost
scientific design and simulation tasks.

CRE announced a new software technology which removes the
need for special-purpose 3D graphics computer chips for many
3D computer applications. This development, which can be
thought of as a software "graphics engine", allows interactive
3D graphics applications to run on ordinary personal computers
(compatible with those found in almost every office) at a
speed previously associated with specialised 3D scientific
workstations costing up to $100,000.

With ramifications for CAD, scientific visualisation,
business graphics, virtual reality and modelling applications,
the new technology provides a very high speed 3D rendering
engine entirely in software.

The implications for application developers are numerous:

o the potential for 3D applications is hugely expanded since
this technology can now be used to run them on ordinary
business computers;

o the same applications can be transfered to many different
machines within an organisation (the design department's
Unix workstations, the marketing department's Macintoshes,
even the Managing Director's PC) since the software engine
is intrinsically portable;

o 3D applications can run with almost any type of display
since the engine is device independent;

o remote / shared applications allowing multiple users to
interact in 3D across continents using only standard tele-
phone lines is now a reality since the engine provides a
low-bandwidth communication facility;

o education needs can now be addressed, since the cost of
the computers needed for the software is much lower than
before; also, the "shared" nature of the software allows a
teacher to control the displays of all the students at
once.

Adam Billyard, Canon 3D Graphics inventor said:

"Before now, interactive 3D performance was only
available on high-end 3D workstations, or by adding
expensive special-purpose 3D graphics hardware to ordinary
workstations. What we did was to throw the textbook out the
window, and approach the problem of providing interactive 3D
graphics performance from a completely new perspective. The
result is a software graphics engine which performs at very
high speed on ordinary workstations and high-end personal
computers."

Dr. David Lau-Kee, 3D Graphics project manager said:

"We are all very excited about the possibilities for
this technology. Already we are co-operating with a number
of sites around the world to refine the technology into its
most useful form. The problem we now face is not `where can
we apply this', but `where do we apply it first'".

Peak performance figures for the software (measured for
typical models in triangles being locally and perspective
transformed, clipped, flat-shaded rendered and software
double-buffered to the screen) are as follows:

IBM PC (486/33) 40000 polygons / second

Sun SPARCstation 1 45000 polygons / second

Sun SPARCstation 2 80000 polygons / second

Performance figures for the Macintosh and NeXT computer
are not available at this stage.

The revolutionary design of the software also leads to
an added bonus: it lends itself well to "distributed" 3D
applications on networks of PCs. At Rennes, a demonstration
of this advance is provided when a 3D Graphics "virtual
office" is co-designed on 4 computers simultaneously. Users
on different computers each have a view of the same 3D scene,
which they can interactively move around. When one user
moves say a "virtual desk", all the other users instantly see
the same changes from their viewpoint. Another demonstration
shows the company logo spinning in 3-space from computer to
computer. The surprising things here are that: first the
demonstrations occur between radically different machines
(ordinary business PCs and Unix workstations); second,
despite the visual complexity of the applications, the
amount of data communication required between the computers
is tiny... so small that it could easily be carried between
continents by conventional telephone lines with little or no
degradation in performance.

Canon Research Europe (CRE) was established on the
Surrey Research Park in Guildford, UK, in 1989. Already, one
of CRE's inventions, the "Wide Imaging Stereo" loudpeaker,
is being marketed throughout Europe after being developed
into the S-50 loudspeaker by Canon Audio, Woking, UK.

The 3D graphics research began in early 1991 when a
different research project required interactive 3D
animation. Since the researchers did not have expensive,
special purpose 3D workstations available for use, they
decided to see if they could devise alternative approaches
which would let them proceed with their work using just
ordinary workstations and personal computers. The
breakthrough came in June last year. Dr. David Lau-Kee
recalls: "We were having a design meeting to discuss what
appeared to be a fundamental barrier to our approach, then
suddenly we came up with the answer. It was one of those very
rare moments in research when it is as if the light-bulb has
just been switched on!"


- END -

-----------------------------------------------------

So, I cant wait for my 3-D CAD-NeXTStation for $5k !!!


Barry Merriman
UCLA Dept. of Math
UCLA Inst. for Fusion and Plasma Research
ba...@math.ucla.edu (Internet; NeXTMail is welcome)


0 new messages