Introduction - Computer Animation: Algorithms and Techniques - page 23

produced by opening the shutter of the film camera and essentially scan converting the elements (e.g.,
polygons) by drawing closely spaced horizontal vectors to fill the figure; after scan conversion was
completed, the shutter was closed to terminate the image recording. The intensity of the image could
be regulated by using the intensity control of the vector display or by controlling other aspects of the
image recording such as by varying the density of the vectors. An image of a single color was generated
by placing a colored filter in front of the camera lens. A full-color image could be produced by breaking
the image into its red, green, and blue components and triple exposing the film with each exposure
using the corresponding colored filter. This same approach could be used to produce animation as long
as the motion camera was capable of single-frame recording. Single-frame recording required precise
frame registration, usually available only in expensive film equipment. Animated sequences could be
colored by triple exposing the entire film. The programmer (animator) was fortunate if both the camera
and the filters could be controlled by computer.
The earliest research in computer graphics and animation occurred at the Massachusetts Institute of
Technology in 1963 when Ivan Sutherland developed an interactive constraint satisfaction system on a
vector refresh display [ 41 ] . The user could construct an assembly of lines by specifying constraints
between the various graphical elements. If one of the graphical elements moved, the system calculated
the reaction of other elements to this manipulation based on satisfying the specified constraints. By
interactively manipulating one of the graphical elements, the user could produce complex motion
in the rest of the assembly. Later, at the University of Utah, Sutherland helped David Evans establish
the first significant research program in computer graphics and animation.
As early as the early 1960s, computer animation was produced as artistic expression. The early
artistic animators in this period included Ken Knowlton, Lillian Schwartz, S. Van Der Beek, John
Whitney, Sr., and A. M. Noll. Typical artistic animations consisted of animated abstract line drawings
displayed on vector refresh displays. Chuck Csuri, an artist at Ohio State University, produced pieces
such as Hummingbird (1967) that were more representational.
In the early 1970s, computer animation in university research labs became more widespread. Com-
puter graphics, as well as computer animation, received an important impetus through government
funding at the University of Utah [ 14 ]. As a result, Utah produced several groundbreaking works in
animation: an animated hand and face by Ed Catmull (Hand/Face, 1972), a walking and talking human
figure by Barry Wessler ( Not Just Reality , 1973), and a talking face by Fred Parke ( Talking Face ,
1974). Although the imagery was extremely primitive by today's standards, the presentations of
lip-synced facial animation and linked-appendage figure animation were impressive demonstrations
well ahead of their time.
In 1972, Chuck Csuri founded the Computer Graphics Research Group (CGRG) at Ohio State with
the focus of bringing computer technology to bear on creating animation [ 10 ] . Tom DeFanti produced
the Graphics Symbiosis System (GRASS) in the early 1970s that scripted interactive control of ani-
mated objects on a vector display device. Later in the 1970s, CGRG produced animations using a
real-time video playback system developed at North Carolina State University under the direction
of John Staudhammer. Software developed at CGRG compressed frames of animation and stored them
to disk. During playback, the compressed digital frames were retrieved from the disk and piped to the
special-purpose hardware, which took the digital information, decompressed it on the fly, and con-
verted it into a video signal for display on a standard television. The animation was driven by the
ANIMA II language [ 15 ] . In the mid-1980s, Julian Gomez developed TWIXT [ 43 ], a track-based
key-frame animation system.