Vicious Circle: John Whitney and the Military Origins of Early CGI

John Whitney himself envisioned nothing more and nothing less for the future of CGI than a union of space and time: “Time has become visual.” But all film deals in the visualization of time. Does this mean that Whitney meant CGI to be the apotheosis of film? Given the brain-numbing use to which CGI is often put in present blockbuster capacities, this possibility is cause for consternation. Perhaps Whitney rather meant the computerized union of color and time as a limit: CGI forms whole worlds beyond what practical effects can begin to mimic, but the technology is only revolutionary where it refuses to supplant practical effects for the sake of mere convenience.

* * *

Alfred Hitchcock’s Vertigo (1958) was the first feature film to use computer animation. John Whitney’s Catalog (1961) was the first computer-animated experimental film. The common thread between these works is Whitney himself, the forefather of CGI. This survey of Whitney’s career posits that CGI technology is inextricable from war technology, and that the psychedelic uses to which early CGI was put are in turn inextricable from this technology. In the current era, when CGI aims less often to convey the impossible than the hyperreal, the military origins of Whitney’s innovations become a matter not only of what movies can do, but of what they should do.

Whitney’s training was not in film but music. From 1937 to 1938, he studied twelve-tone composition in Paris, which eschews keys by using all twelve notes of the chromatic scale and emphasizing none. During the Second World War, as the Americans and British developed early computers, so-called “Turing machines,” to break Nazi codes, Whitney photographed high-speed missiles.1 He realized that the trajectories calculated by bomb site and anti-aircraft gun targeters could be plotted to produce graphic representations of musical harmony. These visuals could then be traced to create motion that reproduced this harmony in space and time. Like the musical harmonic notations that composer John Cage began to develop around the same time, Whitney’s notations prioritized the rhythmic structure of music; the graphically plotted animation of a piece of music would morph according to the melody’s duration.

In 1948, Whitney was awarded a Guggenheim Fellowship that allowed him to develop this method of mechanical animation. He used it in the Five Film Exercises (1940-45) he made with his younger brother James, who became an abstract filmmaker virtuoso. In this film, light is not reflected from drawings (as in traditional animation). Rather, by a proto-slit-scan technique, light itself is carved and stretched through permutations of cut-out masks.

Through the 1950s, Whitney animated TV program sequences, directed animations for United Productions of America, worked as a film specialist for the Charles and Ray Eames Studio, and directed guided-missile project films for Douglas Aircraft. On the basis of this work, Hitchcock hired him to animate the opening spiral sequence of Vertigo in collaboration with graphic designer Saul Bass – who would later use another UPA alum, Peanuts animator Bill Melendez, on his title sequence for It’s a Mad, Mad, Mad, Mad World (Stanley Kramer, 1963).2

For the Vertigo sequence, Bass drew spiraling, twisting shapes based on graphs of parametric equations by the 19th-century mathematician Jules Lissajous. However, Whitney knew that no traditional animation stand could modulate this motion without tangling its wiring. To plot it, Whitney used a “cam machine” made from a rotating M-5 anti-aircraft targeter, a surplus from the war. It had 11,000 components, weighed 850 pounds, and required five soldiers to run, but it was a computer. Each user would put in one variable, e.g., velocity or altitude. Whitney connected the electrical outputs to servos – devices that controlled target positioning – and programmed the targets to move in mathematically controlled ways.3 He called the movement “incremental drift.”4

He then mounted the M-5 and his animation celluloid on a platform, with a pendulum above it. From the pendulum, he hung a pen connected to a 24-foot-high pressurized paint reservoir. As the M-5 moved in tandem with the pendulum, it drew the spirals that open Vertigo. Credit text and trembling, sepia-tint technicolor closeups of Kim Novak whirr and invert as gem-toned spirographs. As in the phenomenon of vertigo, the sense of rotation is a false one. Through this effect, Whitney pioneered the motion-control photography technique still widely used in special effects today.5 Just as modern motion-control photography involves the combining into a single shot of several shots filmed using the same camera motion, Whitney’s technique involved the animation into a single graphic of several targets plotted as they moved along the same mathematically formulated route.

The metamorphoses of text and still and moving images enabled by motion control and military equipment did wonders for Whitney’s commercial career. Over the next few years, he worked with Bass on TV graphics for Dinah Shore and Bob Hope. In 1960, Whitney founded his own company, Motion Graphics Incorporated,6 which soon released Catalog (1961), a seven-minute demo reel of the work he made with his jerry-rigged war surplus machine. The film features basic geometric forms and Lissajous curves multiplied, disintegrated, twisted, and stretched into permutations and mandalas writhing to free jazz by Ornette Coleman.

Catalog swiftly became a classic of the psychedelic age, inspiring special-effects pioneer Douglas Trumbull to recreate Whitney’s slit-scan technique for the light-distorting “stargate” sequence of Stanley Kubrick’s 2001: A Space Odyssey (1968). Ironically, Whitney’s own proposal for the film, a CGI monolith, was rejected. Though Trumbull’s ingeniously disintegrative effect was relatively simple – achieved by moving the camera toward a slit behind which colored, rig-mounted transparency sheets glided – it influenced cinematic depictions of space and psychedelic experience through the present day.

The earliest heir of Trumbull’s pixel effect was Michael Crichton’s low-budget sci-fi Western Westworld (1973). This was the first feature film to use CGI not merely in the title section, but also during live-action scenes. In the film, Yul Brynner plays a robot gunslinger who kills John Blane (James Brolin), a visitor to Westworld, an Old West section of a high-tech amusement park. In the digital scene, the gunslinger stalks Blane’s friend, Peter Martin (Richard Benjamin).

Crichton wanted the scene to portray what his script described as the Gunslinger’s “bizarre, computerized image of the world,” with “flashed-up calculated figures” and “shifting green tones which apparently represent shifts in the Gunslinger’s concentration.”7 After a prohibitively high quote from the Jet Propulsion Laboratory in Pasadena, which estimated against MGM’s $1.25 million budget that two minutes of computer-generated footage would take nine months and $200,000 to create, Crichton turned to John Whitney.

Whitney introduced Crichton to his burgeoning filmmaker son, John Jr., who had each frame of raw footage color-separated into pixelated squares. The effect’s intention – the Gunslinger’s computerized targeting – eerily paralleled the military origins of Whitney’s animation technique. This involvement in CGI and countercultural film did not end with either John: John Sr.’s other son Michael had made his own kaleidoscopic CGI film, Binary Bit Patterns (1969), while his third son, Mark, later directed a relatively sober Carl Jung documentary called Matter of Heart (1986).

While little is known of Michael beyond his sole work, Mark Whitney’s interest in visual art reflected his father’s aims. The abstract expressionist painter Sam Francis, about whom Mark later made a short documentary (A Portrait of the Self, 2007), funded the Jung documentary on the condition that Mark would direct it. Mark, in turn, was enthralled by Francis’ “cybernetic” painting method: rather than painting along a geometrically predetermined course analogous to John Sr.’s CGI technique, the artist – a devout Jungian – entered a trancelike dream state and added circular pools and drips to grids of paint applied with a roller, until geometric regularity arose from the resulting matrix.8 Given that Mark first bonded with Francis over their shared interest in the visual representation of cybernetic formulas, it’s likely that Mark shared his father’s CGI work with the painter.

Regardless, movies through the rest of the ’70s largely failed to catch on to John Sr.’s vision for the revolutionary potential of CGI. The Westworld sequel Futureworld (Richard T. Heffron, 1976) featured a rotating 3D model of actor Peter Fonda’s head, also created by John Jr. in collaboration with fellow CGI creator Gary Demos. The film also included a rotating 3D hand created by Ed Catmull, a computer science graduate student who in 1986 co-founded Pixar. Star Wars (George Lucas, 1977) had a wireframe-rendered trench attack plan for the rebel pilots that was computer-animated by CGI artist Larry Cuba.

The sci-fi film Black Hole (Gary Nelson, 1979) was then not only the most expensive film produced by Disney, but the first to attain a titillating PG rating. Black Hole’s green and black fractal grid credit sequence was the longest computer animation in a feature film at the time. Relative to CGI’s present uses in film, John Whitney’s immediate legacy was, in short, paltry stuff. This is largely because until Catmull’s breakthrough in Futureworld, CGI was generally regarded as a labor-intensive, clunky, outsourced, and restrictively expensive process of 2D rendering.

The rotating hand clips in Futureworld derive from a one-minute CGI short Catmull made seven years earlier, as a demonstration of his University of Utah PhD thesis: A Computer Animated Hand (1972).9 His breakthrough was less in technique – he outlined the hand by mapping interlocking triangles and polygons on a wireframe model and textured it with light and shadow as other early animators had done, albeit usually less laboriously – than in content. Before Catmull, computer graphics were generally a matter of shapes and vectors. The use of this geometry to realistically render something capable of such an infinite array of movement as the human form seemed implausible. After Catmull’s breakthrough, digital animation and live action were not mutually exclusive but codependent. Accordingly, Michael Crichton’s later film Looker (1981) was the first to feature a CG model of a full human body and to use three-dimensionally shaded CGI.

From the 1980s onward, CGI seemed limited less by the animator’s technology than by their imagination: Tron (Steven Lisberger, 1982) used 3D CGI extensively, particularly for its motorcycle sequence. Jurassic Park (Steven Spielberg, 1993) featured unprecedentedly photorealistic CG creatures and a CG face replacement. Toy Story (John Lasseter, 1995) was the first film made entirely from 3D animation, while The Matrix (Lana and Lilly Wachowski, 1999) was the first to use CGI for “bullet time” slow motion. In this century, Avatar (James Cameron, 2009) uniquely featured impressively photorealistic 3D CG characters in an equally photorealistic world.

Today, CGI is so much a matter of box office course that major films that eschew it for practical effects are often lauded or panned as “retro.” Recall the praise and censure around the spate of non-CGI reboots just a few years ago10: Mad Max: Fury Road (George Miller, 2015); Star Wars: The Force Awakens (J. J. Abrams, 2015); Mission: Impossible – Rogue Nation (Christopher McQuarrie, 2015).

With apparently unlimited means of representation, live-action CGI over the past few years – with the exception of children’s and superhero films – has often sought only to convey the real and hyperreal world more seamlessly. On this front, for example, The Irishman (Martin Scorsese, 2019) was distinguished for the elaborate de-aging of its actors, and Cats (Tom Hooper, 2019) for making humans look more realistically like, well, anthropomorphic cats than any film before.

Perhaps the present crossroads of computer animation in major films – hyperrealism and spectacle – are but two sides of the same coin. The matter brings to mind a letter that Larry Cuba had begun writing to George Lucas, before Cuba was hired to animate Star Wars. He thought: “‘What would computer graphics be in the future?’ It would be photorealistic”; the realism itself would be the spectacle: “you wouldn’t be able to distinguish computer graphics from photographs.”11 Cuba never sent the letter. He never even finished writing it. After all, he would have run the risk of talking himself out of a job.

As feature budgets swell and there is ever more at stake for the technology involved, cinematic alternatives to the use of CGI as an aid to realism wane to psychedelically bloated spectacles, such as Three Thousand Years of Longing (George Miller, 2022) or Everything Everywhere All at Once (Daniel Kwan and Daniel Scheinert, 2022). In their own ironic way, visual effects in blockbusters like these now recall the hallucinogenic origins of CGI more closely than other major films.

Computer animation since then has developed to such a pitch that the only territory that remains largely untreaded, perhaps, is an ethical one. When the real is practically indistinguishable from the digital, one crosses a line that legality fails to trace in, say, reviving a dead actress by superimposing her face onto a body double. When film is so immersive as to supplant the viewer’s reality (as virtual reality intends to do), how porous is the line between the viewer’s choice and the creator’s influence?

John Whitney himself envisioned nothing more and nothing less for the future of CGI than a union of space and time: “Time has become visual.”12 But all film deals in the visualization of time. Does this mean that Whitney meant CGI to be the apotheosis of film? Given the brain-numbing use to which CGI is often put in present blockbuster capacities, this possibility is cause for consternation. Perhaps Whitney rather meant the computerized union of color and time as a limit: CGI forms whole worlds beyond what practical effects can begin to mimic, but the technology is only revolutionary where it refuses to supplant practical effects for the sake of mere convenience. Larry Cuba wrote of the line that separates computer graphics from photographs as a good, not an evil.

In the years before his death in 1995, Whitney returned to his musical roots and made films that matched harmonies with colors, designs “played against action-for-action.”13 His first work on this front was a film called Spirals, made in 1988 and frequently reworked thereafter. As in the 1940s and ’50s, when Whitney had conceived of action as a form of harmony, he now conceived of harmony as a form of action. He wrote: “Fluid, orderly action generates or resolves tensions much in the manner that orderly sequences of resonant tonal harmony have an impact on emotion and feeling.”14 The line between computer graphics and photographs remained, but the distinction was no longer a matter of realistic and psychedelic images. It was a matter of the effect that these images had on the viewer. Perhaps that’s all: the more of an impact on the viewer’s emotion CGI has, the more those movies would do well to embrace and expand one’s ability to feel.

* * *

All images are screenshots from the films discussed.