COMPUTER ART
asa

AY OF LIFE

By Gene Youngblood

hen I think of computer art I think of
Chicago, and of four people there—
Jane Veeder, Phil Morton, Dan Sandin
and Tom De Fanti—who are pioneer-
ing a visual art form that ultimately
is not visual at all, but rather the cre-
ation of language, and of conversational ‘‘envi-
ronments” out of which will emerge our future
images and the images of our future. Together
with Steina and Woody Vasulka in Santa Fe,
they inspired me, taught me, and changed me
profoundly over the last ten years. With infinite
patience and dedication, with the passion of vi-
sionary seekers, they guided me ever deeper
into the digital domain and caused me at last to
understand that computer graphics is some-
thing more than art, that it is a kind of practical
philosophy, a way of life, a way of being in the
world and a way of creating a world to be in.

Nostory reflects this more vividly than that of
Jane Veeder, an artist-programmer whose life
and work epitomize both the unique computer-
art community in Chicago and a personal path-
way of growth and discovery that will become
representative of the life of the artist in our
time. In my opinion, Veeder stands with Ed Em-
shwiller and Larry Cuba as one of the most
gifted computer artists working in America to-
day. Relatively unknown until recently, she is
beginning to get the recognition she deserves.
Her 1982 animation Montana is the only work
of computer graphics in the Museum of Modern
Art’s video collection, and her interactive paint
program/arcade game Warpitout—the sensation
of the SIGGRAPH ’82 Art Show in Boston—will
later this year be installed at the Ontario Sci-
ence Center in Toronto, one of the most presti-
gious science museums in the world.

Gene Youngblood was a member of the jury for the SIG-
GRAPH ’83 Exhibition of Computer Art. He is working
on a new book, The New Renaissance: The Computer
Revolution and the Arts.

oto cred p Dodswo

26 FALL 83

“The computer is the future. ..and
remain autonomous and at the same
connected with the future... ’”’

MONTANA: A constellation of totem-like icons dancing on a dreamworld stage.

The road to success has been a long one for 39-year-old
Veeder, whose personal journey as an artist—involving
excursions into theatre, painting, ceramics, sculpture,
filmmaking and video—reflects a widespread disen-
chantment with traditional art forms, a conviction that
the leading edge of culture is no longer defined by what’s
exhibited in art galleries or purchased by museums.
“The go-nowhereness of traditional art forms just kills
me,’ she said. “First painting. The go-nowhereness of
painting killed me off it. And then ceramics. I did ceram-
ics for ten years and after a while you sort of know every-
thing. You start reinventing the wheel. First there was
abstract ceramics, then sculptural ceramics, then utili-
tarian, then Japanese, then you run out of things and
you go around and around. But with electronic stuff, es-
pecially computers, there’s always so much you don’t
know about. It’s this incredible carrot that just keeps
hanging out there in front of you. The computer is the
future of everything and it’s a tool for building tools: you
can remain autonomous and at the same time literally
create the tools that keep you connected with the future
of the world.’

Veeder’s migration from the analog world of video syn-
thesis into the digital domain of computer synthesis is a
trailblazing journey that more and more electronic art-
ists will follow as the advantages and satisfactions of
computer life are recognized. “I like video but it’s very
expensive, so I couldn’t afford personal tools. Also video
skills don’t get much in the way of rewards. Video’s very
easy to do. Many people have these skills. It’s not very
strategic. But having my own computer graphics sys-
tem has made a huge difference. It’s like doing video
without the videotape recorder. Videotape recorders are
horrible machines. They’re complex, they're expensive,

they don’t last very long. That whole electromechanical

combination is problematic.

“Also, being collaborative, video always involves has-
sles with other people. But with a computer, if the soft-
ware is any good at all, the conversation is incredibly
direct. It’s encouraging to know that if you keep trying

you can make it work. You actually are encouraged to
apply your energy. That’s so satisfying. How many rela-
tionships have you had with humans where you knew
that no matter how hard you tried it just wasn’t going to
get any better? So on one hand I just want to talk to the
computer because we understand each other so well.
And on the other hand it gives me a whole other level on
which to communicate with people. With the computer I
can get at my dreamlike images; I can visualize my
made-up pictures with a lot of power and control and
detail; there’s this detail tweakability that enables me to
get at this weird place; the computer offers more control
in support of that space than video ever did.”

ZGRASS: AN ENVIRONMENT FOR ARTISTS

Veeder works in ZGRASS, a language written by Tom
De Fanti exclusively for real-time interactive motion
graphics. Incorporating the best features of Pascal,
Smalltalk, Lisp and BASIC, ZGRASS is by all accounts
the best graphics language around. Among other things,
it is self-teaching and user-extensible—the users can cre-
ate their own commands and integrate these commands
into the language, expanding and evolving it to fit their
personal style of making images.

ZGRASS runs only on the Datamax UV-1 Graphics
Computer which contains three custom chips designed
for Bally Arcade games like Wizards of War and Gorf—
one chip for processing instructions, one for controlling
the display, and one for sound synthesis. This makes
ZGRASS faster than anything around at its price (about
$10,000), and speed is essential for real-time animation.
There’s also a 16-screen memory that allows you to cre-
ate and store sixteen different full-screen images and
then call them up in a real time animation cycle. Output
is standard NTSC video. Veeder works with a single Sony
5850 videotape recorder and builds her editing com-
mands into her graphics programs. “Every tape can be
first-generation,” she explained. “It’s just as easy to run
the computer and make a first-generation tape as it is to
do a dub of the master. And with the 16-screen memory I

SEND

27

it’s a tool for building tools: you can
time create the tools that keep you

a a es

can work up the individual screens and then program
them sequentially and the animation just happens.”

The UV-1 offers less resolution and fewer colors than
larger, more expensive systems—only 320 x 202 x 2 bits
per pixel (256 colors with four colors per area). This re-
presents a tradeoff in favor of real time operation while
keeping the cost down; but if you're moving images you
don’t need that much resolution anyway; moreover, al-
though Hi-Res has been elevated to a quasi-religion in
the commercial graphics world, the fact is that for many
artists real-time interaction is far more important. ‘“‘Pho-
tographic realism is not my goal,” Veeder explained. “I
can see its utility but I don’t think it would bea triumph,
you know. I have no desire to use a super-high resolution
system because I don’t draw in super-high resolution. I
just go-so far and beyond that it’s the integration and
manipulation that I enjoy—movements and gestures.”

In fact, animation per se is not the primary motivation
for Veeder’s involvement with ZGRASS. “I’m just inter-
ested in real time graphic interaction that results in a
dynamic visual event,” she explained. ‘‘WhatI want is to
integrate my eye, hand and brain with the computer’s
ability to perform complex relationships very rapidly.
The fact is I'm addicted to the high-speed personal evo-
lution and perceptual education you get from continu-
ous contact with a real-time interactive machine intelli-
gence.” Nevertheless, Montana and Veeder’s most
recent computer animation, Floater, rank with the best
works of experimental hand-drawn animation. The
quirky, totemic, dreamlike, cyclical image-events of
Montana are reminiscent of Harry Smith’s Early Ab-
stractions, whereas Floater inspires comparison with
Robert Breer, Paul Glabicki, and the optical-printing
films of Pat O’Neill. Both works are essentially auto-
biographical.

MONTANA
For. five years, Veeder and Phil Morton spent their
summers Camping out and making video in the moun-
tains of the Western Badlands—Montana, Wyoming,

Utah. The trips were part of a unique lifestyle approach
to video, a desire to live with and through the medium
conversationally, incorporating computer graphics into
semi-didactic “simulations” of imagined and desired
video realities. They “processed their life” in the elec-
tronic domain, producing video “communiques” of the
realities they created by living and processing them. The
visual signature of these tapes is absolutely unique,
whith Jane’s poignant, ideosyncratic computer graphic
complementing Phil’s sublime image processing—all of
this superimposed over, and integrated into elegant
black-and-white tapes of animal life and geological
splendors of the American wilderness. The tapes have a
joyous, chatty, conversational, grass-roots flavor that
resonates against the otherworldly strangers-in-a-
strange-land consciousness that pervades them. They
are pioneering works by two genuine pioneers of the
Electronic Life.

Gradually Jane evolved away from video and deeper
into computer graphics. She and Morton no longer work
together, but her love of the wilderness, and much of the
graphic material she evolved for the videotapes, con-
tinue to inform her work in computer animation. This
imagery—soaring hawks, mountain peaks, stampeding
buffalo, erosion patterns in the terrain—are combined in
Montana with icons of the technological world: video
cameras, the Sears Tower in Chicago, the Space Shuttle.
Veeder organized the material into 16-screen sequences
programmed as a loop that increases in speed over eight
cycles, with each of the elements moving at different
speeds. The effect is a constellation of totem-like icons
dancing on a dreamworld stage—hence the similarity to
Harry Smith. Equally remarkable is the stereo
soundtrack, which consists of an audiotape of birdcalls
from the Audobon Society combined with sound synthe-
sis performed on the Sandin Image Processor. The effect
of the narrator numbering and naming the birdcallsina
cascade of electronic echoes perfectly complements the
imagery and gives the tape a poignant, haunting, other-
worldly flavor.

28

FALL 83

“Being able to earn my living by
sitting at home with my intelligent

machine...

control over my
life is immense.”

WARPITOUT an installation
in the form of an arcade game whose tar-
get is everyone's favorite image—their

own, sensation at SIGGRAPH '82.

FLOATER

As impressive as Montana certainly is, the six-minute
Floater (1983) is an even more sophisticated work, a
tour-de-force of subtle animation, a masterpiece of sim-
plicity, nuance and economy of gesture. The complex
canvas of Montana is here reduced to a minimal set of
emblematic objects which serve, in Veeder’s words, “ei-
ther as foils for dynamic graphic processes or manifest a
particular motion behavior intimate to their identity.”
All that is left of the Badlands are two buffalo that gallop
in place and two simple gestures that suggest moun-
tains, water and reflections: squiggly lines and a dia-
mond shape that transforms into a ripple and floats of-

screen. (The ripple, which Veeder refers to as a
‘frequency being,” also represents a sinewave, a natural
phenomenon found not only in the electronic domain
but in water and erosion patterns as well.)

In place of the video camera are two simple icons that
represent Veeder’s evolution into computer graphics: a
rotating constellation of mathematical signs floating
cloudlike in a black void, and two white lines, one verti-
cal, one horizontal, manifesting the x-y coordinates of
the two-dimensional graphic space. A few diagonal grid
patterns constitute the only other imagery, except for
the outline of the artist’s hand that periodically “wipes”
the screen as a transitional device. Floater ends with a
remarkable figure-ground reversal as a grid of black
squares emerges from the black background, obliterat-
ing the image: the ground becomes a figure that is indis-
tinguishable from the ground, and nothing is left.

The lyrical economy of gesture in Floater is strongly
reminiscent of Robert Breer’s Gulls and Buoys, while
the rhythmic clustering of iconographic objects in a
black void is as effective as Paul Glabicki’s Five Improvi-
sations; and the cyclical orchestration of the whole
dance calls to mind Pat O’Neill’s earlier films like Down-
wind and Runs Good. The sound for Floater was created
simultaneously with the animation using the audio-syn-
thesis chip (originally used for sound effects in those:
Bally arcade games) that is resident in the UV-1 system;
in other words, the audio was built into the animation
programs, sharing variables with them. The result is a
beautiful composition of small, sensitively-chosen sonic

the
sense of power and

events (Veeder calls them “skeletal figure-sounds”),
punctuated with intervals of silence, that have a dron-
ing/ringing quality like musical Morse Code. It is a re-
minder that she is as talented a sound composer as she is
a gifted visual artist.

Ultimately, of course, the images and sounds are all
that matter; they are the whole point, the reason why
Veeder is simply a good artist, computer or no computer.
Indeed, art is always independent of the medium
through which it is practised: the domain in which
something is deemed to be art has nothing to do with
how it was produced. The use of the computer in the
creation of animated graphics does not suddenly trans-
form this noble tradition into ‘‘computer art’’—it re-
mains animation, whose status as art will ultimately be
determined by art-historical concerns relevant to anima-
tion, not by any consideration of the computer as a crea-
tive instrument.

WARPITOUT

If there is an art unique to the computer, I suspect it
will not have much to do with producing a drawing or
sculpture or videotape; indeed it might not involve pro-
ducing anything at all—for what is most unique about
the computer is precisely its intelligence, that is. its in-
teractivity. In other words, the great value of the com-
puter is ontological rather than phenomenological—it
has more to do with processes of being in the world (on-
tology) than with the commentaries and judgements we
make as a consequence of being here (aesthetics, phe-
nomenology). This is repeatedly confirmed by computer
artists themselves, whose testimonies are almost always
ontological, seldom aesthetic—always about the proc-
esses of producing art through conversation with the in-
telligent machine rather than about the art itself. This is
what Dan Sandin means when he says that one cannot
understand computer art by looking at it. And it is why
Warpitoutis such a brilliant approach to the idea of com-
puter graphics.

Warpitout is an interactive installation in the form of
an arcade game whose target is everyone’s favorite im-
age—their own. A menu-driven environment of pro-
grams written by Veeder in ZGRASS allows the “player”

Continued on p. 29 >

SAN FRANCISCO

VIDEO FESTIVAL

CATALOG

THE 1983 SAN FRANCISCO

VIDEO FESTIVAL

OCTOBER 6 TO 15

Theatrical use of prop and stage device; manipulation of the narrative
form; computer animation; television production and documentary; and
works expressive of the video medium’s fine arts tradition: this array of
formal approaches promises to free artists’ video from both the iconoclastic
stranglehold of the seventies and the hardware fascination of the past few
years. We may all rejoice.

The diversity of style found in contemporary artists’ video is reflective of
the variety of traditions now influencing production. The recent recognition
of the power and significance of artists’ video has attracted creative workers
from many fields. (Admittedly an awkward admission from a reticent art
world struggling with its own often questionable economic traditions.) Each
group is crafting its own special genre of expression.

The catalog which begins on the next page fully represents this new body
of art, with twenty-four videotapes and collections, installations and perform-
ances:

SEND

29

of Warpitout to make computer art with an image of
their own face, experiencing interactive graphics much
more directly than is possible in the restricted, essen-
tially one-way environment of video games. The system
is housed in a Wizards of War cabinet that contains a 21-
inch TV screen with a black-and-white video camera
mounted above it, aimed at the player. There’s a joystick
and two buttons; push one and your face is digitized.
(“It's one of the few things I’ve ever done that only makes
a still picture.’) Next you select a color map and the im-
age is colorized. Then you are presented with a main
menu consisting of nine options for visually processing
the image. The joystick is used to select a particular op-
tion. Some of them, such as Symbol Draw, Symbol Fill
and Pattern Fill, cause secondary menus to be dis-
played; from these you either select objects to be drawn
as foreground figures, or fill in areas of the image with
symbols or patterns which thus become a background
field. Still other options, like Magnify, Ripple and Edge
Blow modify the image geometrically. For example, the
Ripple option redraws a swatch between two points des-
ignated by the player' using the joystick; the redrawing
process is modulated by a sine wave, rippling the image.

Unfortunately, the arcade game metaphor of Warpi-
tout will remain only a metaphor: it’s too expensive to
manufacture and too slow for commercial environ-
ments. “One-way stuff like arcade games have to be fast
and snappy because one-way is boring,’ Veeder pointed
out. “Also, arcades want to kill the kids off quickly so
they can get the next quarter. But Warpitout doesn’t de-
stroy you—you get to destroy you, and you get to sit
there and watch the computer do it. That’s the joy of the
whole thing. It’s fascinating to watch it all unfold before
you. It gives you this tremendous sense of power.”

AUTONOMY: THE COMPUTER LIFESTYLE

Power is a word used frequently in reference to com-
puters, and for good reason, since a computer is a uni-
versal machine that can contain and become all media,
conferring autonomy upon the user and erasing the dis-
tinction between professional and amateur insofar as
that is determined by the tools to which we have access
as autonomous individuals. Jane Veeder is among a rap-

idly-growing population of computer programmers who,
by working at home with powerful personal tools, exem-
plify the emerging ‘electronic cottage’ lifestyle.

“I never visualized my future in career terms,’ she
said, “you know, jobs. My images were always of pure
lifestyle with no visible means of support. Pure lifestyle
issues. I've never been able to get into the Professional
world for more than a year at a time. I get dissatisfied. I
can’t possibly integrate those junk jobs into-my life. But I
can integrate the computer into my life. I can respect its
power. There’s something to deal with there. Who
wouldn’t want to integrate that kind of power with their
life? I’ve always been into control, you know, and what
better instrument than a computer? By teaming up with
a computer and the ZGRASS language I have finally
found a way to make being Jane Veeder a prime career
slot.”

She earns her living as a freelance graphics designer
doing animated sequences for ads and spots, and asa
programmer, writing programs for ZGRASS graphics
packages. In computer jargon, a package is an environ-
ment, a collection of small programs that mediates be-
tween a user and the software language, in this case,
ZGRASS. “My graphics packages stand between the
user and ZGRASS the way ZGRASS stands between me
and assembler code,” she explained. “So I create envi-
ronments for the ZGRASS user. I turn the computer and
its controls into different things. You have menus and
you choose stuff.” The software modules for Warpitout,
for example, were originally written as graphics pack-
ages for commercial applications.

“Being able to earn my living by sitting at home with
my intelligent machine has made a huge difference,” she
said. “The sense of power and control over my life is
immense. . . I want to evolve so far out there that ] don’t
even know where I was before. I was reading recently
some MIT pundit who was warning of the ‘dangers’ of
using powerful interactive computer languages. One of
the dangers he sees is that you get these feelings of powr.
You start thinking that you're evolving into this sort of
far-out electronic person. Well, the fact is that you are!
And I mean what else do you want to do with your life?

30 -

A
PRIMER

jomputer animation has been in
development and use for years, but
only recently has become a common
feature of our visual world. We see it
in movies, TV commercials and
video games.

There are hundreds of computer
animation techniques just as there
are hundreds of ways to make music.
To help understand these tech-
niques, we can group the computer
animation systems into categories,
just as musical instruments are
grouped into woodwinds, strings,
percussion and brass.

In this article we will discuss seven
generic systems as illustrated in Fig-
ure 1.

All the computer animation sys-
tems that we look at are composed of
at least two component parts: the
computer and a drawing device. The
drawing device makes the images
which are photographed onto film or
videotaped. The computer controls
the drawing process by sending the
device electronic signals, and the art-
ist controls the computer.

There are two kinds of computers:
Analog and Digital. Usually when we
use the word computer, we’re talking
about digital computers. These are
the ones that send out utility bills
and run video games.

Few people have even heard of ana-
log computers, because when we en-
counter one, it’s called something
other than a computer. An electronic
musical synthesizer, for example, is
actually an analog computer.

Since the machine is. not usually
called a computer, calling the work
produced with it computer anima-
tion has led to a confusion between
analog and digital work. Rather than
argue the semantics of the term

FALL 83

‘computer animation, we’ll consider
techniques based on both kinds of
computers.

PART 1—ANALOG SYSTEMS
1. OSCILLOGRAPHICS

Analog computers were being used
by artists to make films as early as
the 1950s. Figure 2 is from a film by
Hy Hirsch. The system used to create
this film consists of an analog com-
puter and an oscilloscope, and the
technique is called oscillographics.

An oscilloscope is a screen and a
beam of electrons which hits the
screen and draws a point of light. The
point of light can be controlled in
three ways. The beam’s position is

* Figure 2

controlled by turning 2 knobs, one
for Horizontal movement and one for
Vertical movement. A third knob
controls the intensity of the beam,
and consequently the brightness of
the point on the screen.

Instead of using knobs, we can at-
tach wires to these functions and
control them by: changing the volt-
age in the wire. The amount of volt-
age flowing in the wire substitutes
for the ‘position’ of the knob, and de-
termines the position of the beam.

omputer
nimation

by Larry Cuba

The position of the beam is analo-
gous to the amount of voltage in the
wire. That’s why this is called an ana-
log system. In digital systems, num-
bers are used to specify the position of
the beam rather than voltage levels.
The pattern of voltage channels in the
wire is called an electronic signal.

So the problem of making draw-
ings on an oscilloscope is a matter of
generating electronic signals. This
can be accomplished with an elec-
tronic music synthesizer. The classic
music synthesizer (Moog, Buchla,
Arp) is an analog computer.

The film or video artist creates ani-
mating images on the scope by
generating signals with the music
synthesizer and using them to con-
trol the movement of the beam. This
process of composing an electronic
signal by patching modules together
with cables is used on all analog sys-
tems. The differences are in how that
signal is used to make an image.

2. IMAGE PROCESSOR

By generating two independent
signals and using them to control the
vertical and horizontal movements,
a pattern can be formed which com-

Figure 3

SEND

31

SHAPE OF
RECORD GROOVE

2 STRENGTH OF
MPGNETIZED AREAS
ON TAPE

VOLTRGE
CAPACITOR
DISPLAY REPRESENTS
ANALOGUE OF REAL
PHYSICAL QUANTITY

Vector-scan
display monitor

Stroke
display

Oscillographics

Figure 1

pletely fills the screen from top to
bottom with horizontal lines. This is
the scanning pattern used in video
and the pattern is called Raster scan.
. A television picture tube is similar
to an oscilloscope except that the
horizontal and vertical motion is
fixed in such a way so that the beam
is constantly scanning out a raster
pattern. If the horizonal and vertical
movements are fixed, then that
leaves only the Intensity signal to
form the image.

Variations in the beam’s intensity
during the scan produces an image
composed of various shades of grey.
The levels of grey in the image range
from black to white and correspond to
the range of possible intensity levels.

If we use signals generated with a
synthesizer to control the intensity,
then variations in the voltage level of
the signal will result in variations in
grey level. This is how a video syn-
thesizer forms an image. The image
in Figure 3 is from Electronic Masks
by Barbara Sykes and was produced
with the Sandin Image Processor
(Figure 4).

Figure 4

SELECTED

ABSTRACT NUMBERS
IN AREAY |

DISPLAY
> l } A REPRESENTS
| STORED ABSTRACT
@uanriTies,

scan

otion-
display Contre!

control

>

3. SCAN PROCESSOR

The next analog system we’ll dis-
cuss is the scan processor. Video art-
ist Woody Vasulka used a scan proc-
essor to produce the image in Figure
5. A scan processor, like the music
synthesizer system, uses an oscillo-
scope-like display.

In a scan processor, the raster is
scanned as in video, but unlike stand-
ard video, the pattern can be modified
by the user. Any of the three signals
forming the raster pattern (vertical
movement, horizontal movement,
and intensity) can be modified. These

Figure 5

modifications come from either syn-
thesized signals or a video camera.

In both of the video synthesizers
we've looked at, image processors
and scan processors, signals could be
generated internally or with video
cameras.

An image processor is designed to
process the image as encoded in the
standard video signal (that is, the in-
tensity signal only). A scan proces-
sor, on the other hand, can actually
alter the scanning pattern itself (that

is, it can modify the horizontal and
vertical movements of the beam).

If the scanning pattern is modified,
then the result is no longer a stand-
ard video signal. (So, it can’t be fed
directly to a TV or tape recorded.)
The results appear on an oscillo-
scope-like display and so scan proc-
essors belong with oscillographics on
our chart (Figure 1). And just like os-
cillographics, to be recorded the
screen must be filmed or rescanned
with a video camera.

If rescanned with a video camera,
then a video signal is created which
can be processed with an image proc-
essor. So you could produce an image
on a scan processor and process it
and colorize it on an image processor.

These two functions are combined
in a single system called the Scani-
mate, which was developed by Lee
Harrison in the late 60’s. Ron Hays
used a Scanimate system to create
his videotape Canon (Figure 6).

Figure 6

REAL TIME VS. CONVENTIONAL
ANIMATION

In the systems covered so far,
we've seen analog computers used to
generate and process electronic sig-
nals. These signals were used to
drive two types of displays: oscillo-
scopes and video monitors. With
each of these two displays we could
see the images moving on the screen
in front of us. This display could then
be filmed or videotaped like a live ac-
tion scene. Should we call this ani-
mation, though?

In any film, what we see as a mov-
ing image is really a series of stills
projected in sequence at great speed:
24 per second. In live action, these
stills were created by the camera
capturing a series of snapshots of the
constantly changing real world at the
same speed as projection: 24 per sec-
ond.

In conventional animation, each of
these stills is a photograph of a draw-
ing. The drawings are placed on an
animation stand and photographed,
one at a time. When projected back
at 24 frames a second, movement is
synthesized. Film is the mechanism

32

which allows us to project at great
speed.drawings which were created
and photographed at slow speed (or
one at a time).

In contrast to conventional anima-
tion, the systems we’ve looked at so
far create and display the sequence
of images at high speed so movement
is visible directly on the display.

This is called Real Time Anima-
tion. The sequence of images is being
produced and displayed at a speed
which produces the illusion of move-
ment. This eliminates the need for
film as the mechanism of animation.
Once a sequence of images is visible
on the screen, then either film or
video can be used to capture it at live
action speed.

VIDEO ANIMATION

In conventional frame-at-a-time
animation, the recording medium
must be capable of recording in a
still-frame mode.

This is easy for a film camera, hard
for a videotape recorder. So anima-
tion has traditionally been made on
film. More recently, single frame vi-
deotape recording has been devel-
oped, making ‘‘video animation”
possible.

But this ‘“‘video animation” has
been confused with work produced
on a video synthesizer, which has
also been called ‘‘video animation.’
In one case, you have video record
single frames, so you can shoot draw-
ings with it, and in the other, you use
video technology to synthesize ani-
mating images in real time.

4. ANALOG MOTION CONTROL

The movement in conventional an-
imation comes from two sources:

1) changes in the image from one
drawing to the next, and

2) slight movements of the stand
between exposures to animate pans
and zooms. These stand movements
can be automated and controlled
with computers, and since these
movements alone can create anima-
tion, the animation stand can be con-
sidered a drawing device. This tech-
nique is called Motion Control.

Analog computers, because of
their speed, have an advantage over
digital computers when it comes to
generating real time animation.

But for motion control, which is
frame-at-a-time animation, you’re
better off with digital computers be-
cause of their many other advan-
tages. And today’s motion control
systems are all digital. Still, the first
motion controlled stand used an ana-
log computer. It was called The Whit-

ney Cam Machine. Motion Control
techniques opened the door to a
whole new world of special effects
magic in Hollywood. When the
movie Star Wars came out in 1977, it
started a science fiction craze in Hol-
lywood with dazzling effects like the
jump into “hyperspace.”

This jump into “hyperspace”’ effect
is called a streak, and was invented by
John Whitney, Sr., who invented mo-
tion control in the 50’s by building the
first computer-controiled animation
stand in his garage with an army sur-
plus analog computer.

The analog computer controls the
movement of the artwork which is
back lit on a light table. The film
camera looks down from above and
photographs the artwork as it moves.
The principles of the Whitney Cam
Machine were copied.in machines
that Doug Trumball built for the
2001 stargate corridor sequence, the
ones Bob Abel built to make TV com-
mercials, and the one John Dykstra
built for Star Wars. Since Star Wars ,
motion control has blossomed into
an industry standard, but now all
systems are controlled with digital
computers. We’ll see them shortly in
Part 2 on Digital Computers.

Figure 8

WHITNEY CAM MACHINE

The control mechanisms have ad-
vanced and become quite elaborate
since motion control was invented,
but the basic vocabulary of effects
was established on the Whitney Cam
Machine:

1) multiple exposures,

2) streaks, and

3) slitscan.

One benefit of motion control is the
precise repeatability of animation
stand movements. This ability can
be used to produce multiple expo-
sures. If, while the shutter is open, an
object that emits light is moved into
several positions and its light turned
on, then multiple images of the ob-
ject will appear on the film.

This principle is applied to anima-
tion in motion control: on an anima-
tion stand, artwork can be lit from

behind. When the artwork is moved
4

FALL 83

around in a circle and the lights
flashed several times during the ex-
posure, a multiple image is created.
This is the technique used by James
Whitney (John’s brother) to produce
the film Lapis. (Figure 7)

Another way to use motion control
is to leave the lights on while the art-
work is moving. This creates a time
exposure effect like a photograph of
car headlights whose movements are
visible as a streak of light. A film, Ter-
minal Self was consisting entirely of
streaks were made by John’s son,
John Whitney, Jr.

This streak effect has often been
mistakenly called slitscan, but
streak and slitscan are two totally dif-
ferent effects. The only thing they
have in common is that they are both

Figure 7

time exposures. Slitscan is a 3rd
technique developed on the Whitney
Cam Machine. In slitscan, the art-
work is not seen by the camera lens
all at once. The artwork is comple-
tely covered except for a sliver: re-
vealed through a narrow vertical slit
in the covering. While the camera
shutter is open for a time exposure
the slit moves across the artwork and
reveals it progressively to the camera
which records an image on film.

If, during the exposure, there is no
movement of the artwork or the cam-
era but only the slit, then the art ap-
pears on the film as it would without
the scanning slit, i.e. normal. How-
ever, if while the slit is scanning
across the artwork the camera zooms
in toward it, then a distortion is intro-
duced: a bulge is produced in the fig-
ure. For example, a rectangle is
transformed into a trapezoid.

These three elements of the basic
vocabulary of the Whitney Cam Ma-
chine can be found in the later (and
current) crop of motion controlled
animation stands that use digital
computers.

PART 2—DIGITAL SYSTEMS

For each of the 4 types of analog
animation systems we have covered,
there was a different KIND of analog
computer. One was a Scan processor,
another was an Image processor, and
soon.

~ SEND

33

We'll look at several digital anima-
tion systems, but only one kind of
digital computer. That’s because ail
digital computers do the same thing
fundamentally: they follow lists of in-
struction and can store these lists
(and other data) internally.

The major variations among digi-
tal computers are:

1) the speed at which they operate,

2) the size of their internal storage.
These are usually a function of the
size, cost, and age of the computer.

What distinguishes the digital ani-
mation systems on our chart in Fig-
ure ] is the type of display. We’ll con-
sider three different kinds (vector,
raster, and animation stands), but
they're not mutually exclusive; a sin-
gle system can include more than
one type of display.

ANALOG VS. DIGITAL

The usual method of distinguish-
ing between analog and. digital com-
puters is by comparing the way they
process signals internally: analog
computers use a gradual scale,
whereas digital computers use dis-
crete on-off elements. But these char-
acteristics are at a microscopic level

Figure 9

and do not directly affect the user. A
more useful distinction is at the level
of user control. The way in which an
instrument’s control is structured
will determine what will be easy to do
and what will be hard.

For example, imagine trying to
drive a car by typing commands ona
keyboard, or trying to write by guid-
ing a pencil with a steering wheel.
Clearly, then, the control structure of
a tool can determine its usefulness to
a large degree.

If we compare the control struc-
tures of analog and digital computers
we find...

Analog computers are controlled
by routing signals through modules
with cables, and by turning dials.
Digital computers can be controlled
in a number of ways, which may in-
clude turning dials of drawing witha
pen, but primarily they’re controlled
by typing in commands at a termi-
nal. A list of these commands is a

computer program. A program is just
a description of the steps required to
complete a task. In the past when
someone might have built a mchine
to perform a particular task, now one
need only describe the process in a
program. A digital computer is a ro-
bot which will exhibit the behavior
specified by the program and so be-
come the machine desired by the
user.

This is the most important feature
of digital computers, and apparently
the least understood by the com-
puter illiterate. For example, when a
TV journalist reports a story with the
sentence, ‘“‘Some students at MIT
have developed a computer that can
play chess,’ they invariably mean,
not that they’ve built a computer but

Figure 10

that they’ve developed a program.
We use the word ‘computer’ to mean
a general purpose programmable
machine. As soon as it’s designed for
a specific purpose, then it’s no longer
a computer. Also, once a program is
stored in a digital computer, you can
use it at any time in the future.

With analog systems, the way in
which the modules are connected is
the ‘program, and there’s no way to
store it. If you take it down so some-
one else can use the machine, you'll
have to put it back together again
next time you want to use it.

And there is another difference be-
tween analog and digital computers:
in the analog systems, we saw that
values, such as how far the beam is to
be moved on the screen or how bright
it is, were encoded in the amount of
voltage in the wires. On digital com-
puters, when using a vocabulary of
commands to control the display,
these values are conveyed with num-
bers. Inside the computer, electronic
signals encode these commands and
numbers, but unlike the situation in
analog systems, the user does. not
have to deal with them as much.

5. DIGITAL MOTION CONTROL

The Analog Cam Machine move-
ments are controlled by an electronic
signal. There is a correspondence be-
tween amount of voltage in signal
and position of artwork being moved.

In the digital system, the anima-
tion stand is structured the same,
but the user doesn’t control its move-
ments with signals per se. Instead,
commands are sent via the computer
with numbers to specify the desired
positions of the artwork and the cam-
era. For example, for a simple streak
the program might look like this:

Open shutter
Zoom IN 50 units
Close shutter

This program puts one streak on
one frame of film. For a stationary
image to appear on film, this streak
must go on every frame. So what we
want to happen is for the computer to
repeat this sequence of commands
once for each frame. A 5 second effect
lasts 120 frames, so we add a Repeat
command, causing this group of
commands to be repeated 120 times:

Repeat 120: [Open shutter
Zoom IN 50 units
Close shutter
Zoom OUT 50 units]

After the frame is exposed, the com-
mand, Zoom OUT 50 units is used to
reposition the camera to begin the

Figure 11

next frame. (There is also an assump-
tion here that opening and closing the
shutter advances the film one frame).
This puts the same streak on each
frame, producing a static image. Fora
moving streak, we introduce a slight
change between frames, such as a
small vertical movement of the art-
work. This we do with the command,
“Move UP 2 units”:

Repeat 120: [Open shutter
Zoom IN 50 units
Close shutter
Zoom OUT 50 units
Move UP 2 units]

So as we zoom in and out to streak
each frame, the artwork slowly
moves up on the stand and the streak
moves up in the frame. Multiple ex-
posures and slitscan are also pro-
duced on digital motion control
stands. Figure 8 is a slitscan image
from Ron Hays’ videotape Tristan
and Isolde.

34

6. VECTOR GRAPHICS

If we look at our chart in Figure 1
again, we see two types of electronic
displays: Raster scan and Vector
scan. The difference between the two
is the way in which the beam scans
the screen to create an image (like
the difference between an oscillo-
scope and a video monitor). Raster
scan, as we learned before with ana-
log systems, is the scanning pattern
that video monitors use. The whole
screen is scanned, one line at a time,
from top to bottom. The image is
formed by varying the intensity of
the beam during the scan.

A vector scan display, on the other
hand, is like an oscilloscope, in that
the image consists of lines (called
vectors) which are drawn by the
beam as it moves around the screen.

But again, in digital displays the
user doesn’t create the signals that
contro] the beam. These signals are
created internally by the display in
response to commands. The com-
mands instruct the display where to
draw points and lines by means of
numerical coordinates.

There are various kinds of vector
scan displays and they can be
grouped into three classes based on
three ways of recording an animation
sequence: film recorders, storage
tubes, and real time displays.

In a typical film camera, like a Bo-
lex, we can identify three ways to
shoot animation: single frame with a
time. exposure, short exposures (still
single framing) and live action speed.
In single frame mode, we can shoot
time exposures by holding the re-
lease down. This holds the shutter
open. The second way is also single
framing, but pushing the release

Figure 12

gives us a short exposure (1/30th
sec.) no matter how long we hold
down the release. And the third way,
live action speed, is 24 frames/sec.

1) time exposure—A film recorder
is a vector-scan display with a cam-
era mounted facing it so a lengthy
drawing process can be captured on
each frame of film as a time exposure
(not unlike the streaking process of
motion control).

Since the drawing accumulates on
the film over time, it can’t be seen in
its entirety by the user until the film
is developed. This isn’t very conven-
ient, so these displays are usually
used in conjunction with another
type of display which can be used for
previewing the image. This type of
display is called a storage tube.

2) short exposure (but still frame-
at-a-time)—the storage tube display
has a screen which retains the image
of the points and lines as they’re be-
ing drawn instead of allowing them
to fade, so that the user can see the
whole drawing all at once and a cam-
era can record the drawing on one
frame of film with a short exposure
(rather than a time exposure).

The process is still one-frame-at-a-
time-animation, though. You can see
individual drawings, but you can’t
see movement until the film is devel-
oped and projected. There are dis-
plays on which you can see move-
ment directly without filming, and
they’re called real time displays.

3) live action speed—Real time dis-
plays draw and redraw images fast
enough to see the animation without
filming or taping (just like the proc-
ess we saw in the analog systems).
Figure 9 is from my film 3/78 which
was animated on a real time system.

Since the vector images are ani-
mating in real time, we can point a
video camera at the display and proc-
ess the signal with an Image Proces-
sor. Figure 10 is from Spiral 5 by Tom
DeFanti and Dan Sandin and was
produced on a system which com-
bines real time vector graphics and
image processing.

SOFTWARE VS. HARDWARE

Regardless of how the image is cap-
tured on film, the programming is a
function of the main computer and
the sofware available for it. The word
‘Software’ has two connotations.
One contrasts Software with the
word ‘Hardware. Hardware consists
of all the pieces of the system which
are material objects, such as the
main computer, the display, the ter-
minal, the printer, the disc drive and
so on. Software, in this context, is the
information that is stored in the vari-

FALL 83

SUNSTON
ous memory devices. Software in-
cludes programs which help you edit
other programs, programs that draw
pictures, or data such as the num-
bers in a phone book.

But in another context, like the
phrase ‘Software Development, the
word software does not refer to data,
but to a class of programs which are
designed specifically as tools. A pro-
gram which helps you edit text, for
instance, is a software tool. Software,
in this sense of the word, can deter-
mine how useful a piece of hardware
will be for a particular task. Software
has a hierarchical nature which al-
lows users to extend the tools pro-
vided, with their own. This is one of
the more powerful aspects of digital
computers and it doesn’t exist in an-
alog. This extendable nature of soft-
ware makes it unique among man’s
tools in that the tool user can also be
the tool builder.

7. RASTER GRAPHICS

Each of the three methods of re-
cording animation (on film or tape)
that we identified for vector graphics
apply to raster graphics as well,
namely:

1) time exposures—the same film
recorder that’s used for vector graph-
ics is capable of operating in a raster
mode, too. As in the vector mode,
each frame in a sequence is stored on
film via a time exposure. Figure 11 is
from a commercial produced by Digi-
tal Effects using a film recorder.

2) short exposure (frame-at-a-
time)—in vector graphics, a storage
tube was used to preview images and
to avoid the need for time exposures
during filming. But there is no such
comparable screen that can store a
raster image. So the problem of dis-
appearing images is solved by creat-
ing a special section of the comput-
er’s memory (called a frame buffer)
for the storage of the data required to
draw one frame. The display screen
is constantly refreshed from the data
in this frame buffer memory. So, as
the main computer builds the image
by changing the numbers stored in
the frame buffer, the results are con-
tinuously visible on the screen.

SEND

35

The resolution of a frame buffer is
: red_in_number_of_nixels .The
Td prsisaconiratuun SS hntrure
Element. The pixels fill the screen
the way squares fill a page of graph
paper. The finer the grid of your
graph paper, the smaller is each pixel
and the more you have. The more
pixels, the higher the resolution.

For each pixel on the screen, there
is a number stored in the frame
buffer which determines the shade of
grey (or color) for that pixel. Chang-
ing the number assigned to a pixel
changes its shade (or color). The
range of numbers is limited and de-
termines how many colors are possi-
ble for any pixel. Since this number
is stored in binary form, its size is
measured in number of bits per
pixel. But you don’t need to know bi-
nary arithmetic to compute the reso-
lution of a frame buffer—just this re-
lationship: 1 bit means 2 colors, 2
bits means 4 colors, 3 bits, 8 colors
and so on. The amount of memory
devoted to each pixel (in number of
bits) is a measure of the color resolu-
tion of the buffer. The complete reso-
lution of the buffer, then, would com-
bine spatial and color, and would be
specified as number of pixels Wide
by number of pixels High by number
of bits per pixel, or Depth. For exam-
ple, the resolution of the frame buffer
in the Datamax Zgrass computer is
320. Wide x 2 bits/pixel. (Figure 12)
High resolution frame buffers were
used by Ed Emshwiller to produce
his videotape, Sunstone, (Figure 13).

3) live action speed—if these buff-
ers can be udpated fast enough, the
animation can be seen in real time.
From video games to flight simula-
tors, real time animation is required
when the images must respond to
human interaction. Stan Vanderbeek
used the flight simulator at NASA‘s
Houston center to create the images
of his film, Euclidean Illusions, in
real time.

The difference in resolution, real-
ism and visual quality between video
games and flight simulators is obvi-
ously enormous. This reflects the
large difference in size and cost of the
computers. But as these factors con-
tinue to drop, who can say how long
it will be before playing a Star Wars
computer game will look like being
IN Star Wars, the movie?

A videotape of this presentation
will soon be available. For more in-
formation write: Larry Cuba, 1803
Mission Street No. 123, Santa Cruz,
CA 95060.

WIeo

Original material needed

por-f6r'mons/
vid é 7 0. n. 1. Emphasis on

consistent strategies for non-commercial experimen-
tation. 2. Productive confrontation with contemp-
orary issues. 3. Fully accredited BFA and MFA
programs.

Applications are now being accepted for the Spring
1984 semester. For a free copy of our current
College Bulletin, with complete information about

all our programs in the fine arts, write or call the

Office of Admissions.

San Francisco Art Institute
800 Chestnut Street

San Francisco, CA 94133
(415) 771-7020

Civa

for ongoing artshow in s f
Clubs/galleries 4156210909