perm filename 02[NSF,BGB]1 blob
sn#073913 filedate 1973-11-29 generic text, type C, neo UTF8
COMMENT ā VALID 00003 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00002 00002 Dave Sneider Stanford University.
C00007 00003 First, the benefits: it is for me to claim (and for you to
C00014 ENDMK
Cā;
Dave Sneider Stanford University.
Diector of Science Software, DACS. Artificial Intelligence Lab
WED Enterprises Stanford, California 94305
1401 Flower Street
Glendale, California 91201
Dear Mr. Sneider,
This letter is essentially a feeler for a grant to support a
basic research program in computer animation. Unlike defense,
transportation and medicine; the government does not fund reseach in
entertainment. Nevertheless, I will try to demonstrate that a novel
situation exists where your particular company can benefit by
influencing the direction of university based computer science
research.
As I am sure you are aware, films with the technical quality
and realism of Snow White will at sometime in the future be generated
by means of script writers and cartoonists using interactive computer
graphics. At that future date, each cartoonist and script writer will
have a display computer console in his office connected to a large
central computer (say two PDP-10's, 300K words core, with 50 mega
words of disk for virtual memory and file storage). The cartoonist
will design actors, objects and scenery by typing in commands to
build completely synthetic objects. An existing physical object (or
actor) could be entered into the computer by means of television
cameras. Script writers would develope their ideas using a script
language similar to their present notations. A "script compiler"
would then compile the script into detailed commands for a "reality
simulator". For developmental work, CRT line drawings and low
resolution video are provided in nearly real time; for final filming,
the computer generates high resolution video images which are
transferred from computer media to celluloid by an FR-80 or
equivalent. The point is that the hardware for making very high
quality, three dimensional, color animation already exists. What is
lacking is the software; and that software does not exist because
scientific problems in three dimensional object representation,
photometric modeling, animation languages, hidden surface
elimination and editing a simulated process have not been solved.
Whether this technology appears sooner rather than later depends on
how much skilled computer science effort is directed at the problem.
Next, I wish to argue that the most expedient, and mutually
advantageous way of developing this technology is for WED Enterprises
to support the relevant computer science research at Stanford
University; there are two points: propriety and benefits.
First, the benefits: it is for me to claim (and for you to
judge) that Stanford can do long range basic exploratory research
that would be too expensive for you to do for yourselves. The
capital cost of our exotic computer system; has been paid for by the
government. The labor cost of highly skilled and motivated students
and research programmers is low. The main results of our work would
be the explicit theory (algorithms, languages, representations and
systems design) required for doing computer animation. We would
advance the state of the art to a point where you could apply it.
Second, my answer to the propriety problem involves
distingushing between "basic research" and "applied research". Basic
research involves creating theories and demonstrating matters of
principle. Applied research involves using the state of the art to
make money. WED enterprises would do its own applied research and
product development which would be secret; and we would do the
computer science which would be in the public domain. Naturely, if
you think that you can do all the research on your own, then it is
to you advantage to keep it a secret and to tell us to "forget it".
On the other hand, if you accept the premise that sophisticated
software can require years of development and that we are as good
programmers as we claim, then it would be to your advantage to have
us begin programming the system for computer hardware that you will
buy in 1976 or 1980. A final minor point is that Stanford University
is a school and will not go into business for itself doing computer
animation. My argument can be summarized in three statements:
1. WED enterprises should obtain by 1980, a large and sophisticated
computer animation system of the size and quality depicted above.
2. The Stanford Artificial Intelligence Project already has the
necessary futuristic hardware and some of the software out of which
such a large computer animation system could be built.
3. The software development of such a system could require three to
ten years of basic research. And the cheapest way to assure success
is to have outside consultants (Stanford), do the preliminary
development and feasibility study.
I have enclosed a slightly reworked draft of a similar
proposal that we were going to submit to NSF (National Science
Foundation); before the energy crisis cut off their funds for
computer science. Also enclosed is a flip book of animation showing a
mechanical arm turning a block over. Finally, I wish to extend an
invitation to you and your associates to visit the A.I. Lab whether
or not you are interested in this present idea. We would welcome
seeing the creators of audio animatronics and would gladly share our
graphics and our robots.
Sincerely
Bruce G. Baumgart, Research Assistant