We know D.C. Get our free newsletter to stay in the know.

Anybody who tells you that 14 hours is a lot of television is either debating whether he can make it through the last episode of Berlin Alexanderplatz or living without TiVo. My first-generation video recorder can cram nearly all of Fassbinder’s magnum opus into its hard drive thanks to an ingenious trade-off between foreground and background: The recorder updates only those parts of the image that have changed since the last scan. Full-screen action shots give the thing fits—rushing water, for example, breaks up into a shimmering silver mosaic—but, lucky for me, the Niagara Channel isn’t one of my favorites.

TiVo’s engineers know that, much of the time, most of a so-called moving image isn’t moving. For a century, this same easily overlooked fact has been responsible for some of cinema’s most enthralling illusions, created through that most ancient of imagemaking technologies, painting. By expanding the scene past the bounds of the set, matte painting has allowed filmmakers to pretend to go where finances and physics won’t permit them to tread.

It’s an ephemeral art. A painting used for a single establishing shot is on screen two to six seconds, never to be seen again. The originals historically haven’t been accorded much respect, glass plates being wiped clean for reuse and Masonite boards often not surviving purges of studio archives. And now, the digital paintbox threatens to make relics of the medium’s brushes and oils; this particular brand of celluloid legerdemain is fast going the way of the magic-lantern slide. Against this backdrop, Cinefex scribe Mark Cotta Vaz and matte painter Craig Barron have teamed up to create The Invisible Art: The Legends of Movie Matte Painting, a history of special-effects innovation and apprenticeship that stretches from the man-with-a-movie-camera loners of film’s infancy to the bureaucratic technocracies that govern FX today.

Having been drawn to the studio of magician and filmmaker Georges Méliès while visiting France as a young man, California photographer Norman Dawn was inspired to take up cinematography. Using his inheritance to buy a state-of-the-art Debrie movie camera and smuggle it back home (the American film companies had essentially formed a cartel), he created a 1907 travelogue of California missions, some of which had fallen into disrepair. He “restored” them by adapting the still-photography technique of positioning a judiciously painted sheet of glass between camera and subject to improve elements of a scene, producing cinema’s first “glass shot.” Later, Dawn would help introduce rear projection, whereby background footage is screened behind an actor as a scene is shot, and pioneer the “original-negative” matte-painting process. This basic technique, whereby black mattes and counter-mattes affixed to the front of the lens allow separate exposures of live action and hand-painted surround to fit together like pieces of a jigsaw puzzle, was, with some variations, to dominate the field until the advent of computers.

Dawn patented his process but wound up in court and eventually was forced to settle, selling the patent for $10,000 to MGM production head and movie-biz 800-pound gorilla Irving Thalberg, who made it available industrywide. This effort to share the technical wealth went only so far, though; inter-studio secrecy meant that each company’s effects shop had to continually reinvent the wheel to keep abreast of the competition. The camera boffins were at least entitled to use what they discovered, but, for decades, they were deprived of credit for their work by bosses who preferred to neither unveil the trickery that powered the dream factories nor admit publicly that it existed.

The craft was advanced by such developments as the optical printer, which allowed multiple filmed elements to be composited within a single frame, and the Dupy Duplicator, which enabled identical camera movements to be replicated in separate shots. For a book with such technical subject matter, The Invisible Art does a lousy job explaining how things work. It isn’t until Page 107, for example, that you actually see a diagram featuring a matte affixed to the front of a camera and labeled as such.

But this coffee-table-ready production from the capable designers at Chronicle excels in providing loads of before-and-after illustrations. Its tiny, blacked-out center awaiting the addition of live-action footage, an Albert Whitlock painting of a devastated Los Angeles is paired with a still from the final scene of 1974’s Earthquake. A shot of the two-story set for Gene Kelly’s building in 1951’s An American in Paris is seen alongside a final composite that incorporates Lou Lichtenfield’s painting of the top floor, complete with a photo of Kelly collaged into an open window. Similar pairings, from nearly 20 movies, are fused together in animations on the accompanying CD-ROM, which runs from the live-action matte composite of a passing train in Edwin Porter’s The Great Train Robbery (1903) to the digital wizardry that animates neon signs in Martin Scorsese’s Casino (1995).

By the time Vaz and Barron tackle the “radiosity” rendering of the latter, they have painted a picture of a trade transformed by technology and taken out of the hands of the artisans who developed it. No master likes to see his craft ceded to upstarts who aren’t trained in the basics because the basics no longer apply. And The Invisible Art’s closing chapters are redolent of the sour regret of the old-timers, many of whom have retired or turned to fine art as the business has switched over to computers.

The book’s account of the crossover is marred by multiple oversights, however, the most glaring of which are all the more puzzling because they involve “legends” singled out elsewhere by Vaz and Barron. Mentioned in the CD-ROM timeline but not the book are such transitional milestones as 1982’s Tron, whose visual effects were supervised by third-generation matte maestro Harrison Ellenshaw, and 1998’s What Dreams May Come, which featured both mattes and a multitude of digitally enhanced effects meant to create the illusion of a painted paradise. Though neither of these pictures comes close to succeeding as an artistic whole, neither does George Lucas’ The Phantom Menace (1999), which makes the book and is not noticeably more important to the history of visual effects.

Perhaps the book’s problem is the Saturday-matinee ideal of transparent realism, as advanced by foreword author Lucas, who haunts the proceedings like a mirror-shot ghost. Although Vaz and Barron note that an early matte by Paul Lasaine for Dick Tracy (1990) didn’t pass muster with director Warren Beatty because it was too realistic, and they concede that the mattes made by Warren Newcombe’s MGM shop for The Wizard of Oz (1939) produced a storybook effect, the overall romantic conception of The Invisible Art requires that the once-secret work of the unsung heroes of Hollywood’s Golden Age still not call attention to itself.

To Vaz and Barron, stylization is salt to illusionism’s meat, not the other way around. Tron was the first movie to matte minimally developed characters into CGI environments for much of its length. What Dreams May Come played out a hash of Dante, the Orpheus myth, and The Bridges of Madison County amid a diarrheal three-dimensional pastiche of Church, Friedrich, Monet, Van Gogh, and Thomas Cole that comes off like a rustic-scene bar mirror animated by Thomas Kinkade. These films are slighted here not because they suck, though, but because their effects are supposed to upstage the story by trumpeting their unreality, in contrast to the book’s identification of matte painting with invisibility.

The authors’ view seems rather limiting if you consider that visual effects ought to be a way to match live-action photography to any conceivable setting, regardless of realism, just as film itself ought to be a means of attaching any sight to any sound, to hell with Hollywood conventions. The continued charm of A Trip to the Moon (1902) is due not to any success in duping us into thinking Méliès beat NASA by 60-odd years but to its dashingly fantastical, and evident, style; to see its effects as the blunt theatrical stunts they are is only to fall deeper under their spell. Vaz and Barron also sidestep the fact that matte painting thrived during its ’30s-to-’50s heyday because Tinseltown hegemony had prevailed in establishing an idiom of outrageously contrived visual styles. Studio-era Hollywood was, of course, a painted world, from the models and backdrops to the sets to the faces of the actors. The admittedly amazing matte work done by Mario Larrinaga, Fitch Fulton, and Chesley Bonestell for 1941’s Citizen Kane convinces because its deliberately curtailed approximation of actuality meets the manifest fakery of the production design halfway. Matte painting isn’t about matching art to life but art to film.

No picture demonstrates this relationship better than The Wizard of Oz, whose visual world was intended to be bought into exactly as would that of a picture book, where a drawing’s looking like a drawing is no impediment to its conjuring fantasy. At the start of the chase scene along the parapets of the witch’s lair, the Scarecrow leads the fleeing gang of good guys around a pillar. In the hubbub, the Tin Woodman and the Cowardly Lion jostle its side, which gives a fabric flap or two before the cut. Decades before “two-and-a-half-D” would be used to describe the CGI application of 2-D texture maps to 3-D wire-frames, MGM’s set designers were performing a real-world version of the same thing, stretching painted canvas around wooden supports. The scene doesn’t suffer, because your investment in the picture doesn’t reside in anything so flimsy as realism. You expect all of Oz to be a bewitching sham rigged with wires and levers; did someone say something about a wizard?

Visual effects succeed not when they disappear but when they are consistent with the overall look and feel of the film and come in support of a solid consideration of the way a particular virtual environment works. The makers of Toy Story (1995), the first commercially viable fully computer-animated feature, knew that rendering convincing human beings was beyond them, so they built their manic quest around characters who naturally would have machined surfaces. Surroundings polished to the same mathematical sheen were used to bridge the divide between the playroom and the world outside. When such matching is infeasible, as in Tron and What Dreams May Come, it is possible to make the textural imbalance a deliberate part of a metaphysical fish-out-of-water story; that these pictures still don’t work can be ascribed to an underthinking and overloading, respectively, of the interaction between character and universe. Lucas’ recent CGI fests, however, succumb entirely to their insistence that flesh-and-blood actors occupy the same corporeal plane as the shiny figments of the computer’s imagination. (It doesn’t help that he has engaged the green-screen talents of Natalie Portman and Jake Lloyd, the worst child actor in this galaxy or any other.)

Having gotten his start in the business as a matte-painting photographer with Lucas’ Industrial Light + Magic, on The Empire Strikes Back, Barron is indebted to Lucas, as is Vaz, who has published a slew of cheesy authorized Lucasiana. So naturally they neglect to condemn the wave of special effects that has swept Hollywood from storytelling to not-quite-convincing spectacle in the quarter-century since Star Wars. For all mattework’s impermanence, it has helped engender an eye-candy aesthetic that has tyrannized the multiplex for decades. The hidden theme of The Invisible Art is that whereas matte painting once extended the creative possibilities of film, on the eve of its obsolescence it can be accused of having played a pivotal role in broadening the cinematic canvas to the point that almost nobody knows how to fill it up. CP