In the VFX ABC, the letter “O” stands for “Optical Printer”.
So what is an optical printer? Simply put, it’s a piece of equipment designed to copy motion picture film. A typical optical printer has a projector at one end and a camera at the other. By running film through the projector, and rephotographing it with the camera, you can create a near-perfect duplicate of your original.
But that’s just the start. By moving the camera and projector around, or zooming their lenses in and out, you can change the size and orientation of the final image. You can alter the timings to slow things down or speed them up. You can employ filters to change colours or introduce blurs. Through the clever use of masks, you can create animated transitions between shots, or combine lots of separate images into one. In short, with an optical printer, you can create almost any visual effect you like.
Many histories of optical printing begin in the mid-1940s, when the Acme-Dunn optical printer hit the market as the first mass-produced device capable of doing all the above (and more). In reality, the Acme-Dunn machine was the culmination of many years of research, innovation and experimentation by a host of dedicated cinematography and visual effects pioneers.
Diagram by Roscoe C. Hubbard illustrating the difference between early contact and optical printers. Image source: Journal of the Society of Motion Picture Engineers, February 1927.
Long before television — let alone DVDs, Blu-rays or streaming services — rudimentary optical printers were being used to service the growing demand for a home cinema experience. In 1918, Moving Picture World described one such machine as “an ingenious device for printing from a standard size motion picture negative on to a smaller film used by one of the many small home projecting machines upon the market1.”
An early optical printer. Photograph: Moving Picture World, 23 March 1918
These basic devices were not exactly commonplace. Nor did they conform to any kind of standard design. Nevertheless, many film enthusiasts had already identified them as the shape of things to come. In May 1922, addressing the Society of Motion Picture Engineers, cinematographer Alvin Wyckoff bewailed the general lack of progress being made in developing a sophisticated and reliable optical printer for professional use.
“There are several good step printers and one continuous printer on the market,” Wyckoff said, “but to my mind they are antique. We should be able to do with our printing machine even more than we can do with our camera. It should be so flexible that we can take any part of our positive and make a new negative; that will enable us to do trick work that as yet has not been seen. There is nothing of this kind as yet on the market.2”
In fact, such machines did already exist, although in severely limited numbers. Responding to Wyckoff’s plea, Pathescope’s Willard B. Cook updated the Society on the availability of optical printers for research purposes: “Mr. A. F. Victor built twelve of them, which are distributed among different laboratories over the country.3”
Optical printers gained ground through the 1920s, performing such mundane duties as copying original negatives, as well as resizing them from 35mm to 16mm, and vice versa. These early printers often used daylight to expose the film, and their output was generally destined for the educational or non-theatrical market. At the same time, enthusiastic inventors continued to explore the creative possibilities of the technology for the rapidly-expanding theatrical motion picture industry.
Press advertisement from 1928 for a Depue-Vance daylight optical printer.
In 1927, A. B. Hitchins of Duplex Motion Picture Industries Inc. presented his latest optical printer to the Society of Motion Picture Engineers. As well as enlarging, reducing, and handling routine optical effects such as lap dissolves and wipes, the Duplex optical printer also boasted a special effects attachment comprising a matte box, multiple exposure device, circular and rectangular vignette, curtain shutter, and blade cut-out.
“The Duplex optical and reduction printer is the result of an insistent demand for improved and more flexible printing methods,” Hitchins stated. “Directly we enter the field of optical printing, we open up a practically unlimited range of printing possibilities; every phase of trick and effect photography can be readily accomplished, limited only by the ingenuity of the operator. With a machine of this type at one’s command many expensive sets need never be built, for any desired detail or background can be printed in by double exposure or with silhouette negatives.4”
The 1927 Duplex optical printer Type A featured a built-in special effects attachment and matte box. Photograph: Journal of the Society of Motion Picture Engineers, September 1927.
The Society benefited from further optical printing insights the following year, when cinematographer Carl Louis Gregory presented “an optical printer for trick work”, designed by himself and built by Fred A. Barber.
Gregory illustrated the challenges of “trick work” by describing the process of putting a mermaid on to the screen. His proposed solution (requiring an optical printer, of course) required the filmmaker to “make negatives of a nude woman swimmer and of a shark or a minnow and from these … dissect the trunk of the woman and the tail of the fish and assemble a mermaid that will almost make you believe in the existence of the fabled creatures.5”
Critical to the success of this operation was the ability to perform the final assembly with great accuracy. “The solution of this difficult problem,” Gregory advised, “is in an optical printer where every mechanical move can be controlled with micrometric precision.6”
Mounted on a six-foot lathe bed on a concrete foundation, the Gregory-Barber optical printer was a “three-head” device. This meant it carried three key optical components — camera, lens and projector — each mounted on its own moving carriage. The carriages, or “heads”, were capable of independent movement up, down, left and right, in increments of one eight-hundredth of an inch.
The Gregory-Barber optical printer was a precision-built optical printer capable of head movement to within a tolerance of one eight-hundredth of an inch. Photograph: Journal of the Society of Motion Picture Engineers, April 1928.
Fixing it in Post
By the 1930s, optical printers were essential tools relied upon by all the major motion picture studios. Not only did they offer boundless creative possibilities, but they also made financial sense by saving time on repetitive tasks. In 1931, RKO’s head of special effects Lloyd Knechtel said, “Within the past few years the motion picture has almost universally adopted the system of optical printing in its many and varied uses and forms, and has found it invaluable as both an artistic and an economic aid.7”
Much of the work running through the gate of a typical optical printer was drudgery — fade-ins and fade-outs, dissolves and all the countless wipes and iris effects familiar to early cinema audiences. However, filmmakers were also seeing in the optical printer the potential for correcting errors made during principal photography — what is now commonly called “fixing it in post”.
“There are ever increasing demands made on the optical printing experts,” Knechtel observed, “to ‘doctor up’ scenes that have already been photographed and which require certain ‘adjustments’ in order to make them usable in a production.8” Optical fixes might include correcting the accidental under or overexposure of a scene, reducing contrast, using coloured filters to create “day for night” effects, or blurring out unwanted details such as a rogue company name on an advertising billboard.
The editor of International Photographer, commenting on an article by Maurice Hughes of Pacific Title and Art Studio, likened optical printer technicians to combat medics: “It is [Maurice Hughes’] belief that the optical printer has been the emergency doctor in many successful operations performed on film9.”
Trick work — what is now called visual effects — enjoyed a boom period during the 1930s, as films ranging from King Kong (1933) to a seemingly endless stream of Busby Berkeley musicals conjured optical illusions to dazzle audiences seeking escape from economic gloom and political upheaval. It was in the field of trick work that optical printers came into their own, not only by stirring audiences with fabulous spectacles, but also by helping producers to cut costs.
In 1934, cinematographer Linwood Dunn noted, “As studio executives become more familiar with the artistic and money-saving possibilities [of optical printing], I feel assured that they will take more and more interest in this branch of trick photography10.” A few years later, in 1937, J. A. Norling identified the economic benefits of the optical printer as the key to its ongoing success: “There is no better way to insert a background in composite photography than by background projection or by the Dunning process, but some composite scenes requiring background insertion can be made on the optical printer at less expense, if the picture action is suitable11.”
The sequence of images below breaks down the different elements used to create a typical visual effects shot of the period. It was created for the film So Ends the Night by Cosgrove Special Effects Department at RKO Pathé Studios, Culver City, under the supervision of Jack Cosgrove12.
For this RKO composite shot, the original live action (top left) was shot on a stage at Universal. The man and dog were further isolated by what would later be known as a “garbage matte” (top right). This was achieved by projecting the original footage on to a white layout board, drawing in the rest of the silhouette by hand, and rephotographing the result. Cosgrove Special Effects combined the result with a stock shot of a sunset sky (bottom left) using an optical printer, with the final composite image (bottom right) achieved after much experimentation with different light levels and a range of high contrast films. Image source: International Photographer, December 1940.
The Philosopher’s Stone
Indispensable though the equipment had become, even by 1940 there was still no such thing as a “standard” optical printer. Individual machines were meticulously hand-crafted, with each studio’s special effects department pioneering its own variations on the theme. As the benefits of optical printing became increasingly apparent, the race was on to mass-produce a truly reliable, standardised device.
One man hoping to be the first past the post was William Draper, who in 1940 described his frustration in trying to get one of his ever-evolving prototypes into production. “With the issuance of letters of patents in 1938,” Draper remarked, “[I] determined to make one grand effort to put [my] system to work. Believe it or not the effort is still in the process of incubation. It seems incredible that such a simple idea would present such a multitude of problems13.”
William Draper’s optical printer of 1940, configured with a travelling matte in the camera head. Photograph: International Photographer, April 1940.
Driven by a desire to improve both productivity and reliability, Draper dreamed of an optical printer that would “reduce thirty minutes of a skilled workman’s time to five minutes of a printer’s time” and give filmmakers “a more standardised, dependable product … as near foolproof as possible14.” However, the complexity of the challenge seemed insurmountable. Working with business associate J. W. Fitts, Draper tinkered endlessly with “an optical printer with a few extra gadgets [which] reminds one of the perpetual motion machine that was constructed by the student of theosophy in the Philosopher’s Stone. When it was contained in a cigar box it lacked one gear of being ready to function. When it had outgrown the house it still was short one gear. No matter what is added, it still suggests the addition of something else15.”
“Here are the details of the ‘perfect’ optical printer that would be the answer to the prayers of motion picture camera operators for a mechanically foolproof device that even eliminate tests” – William Draper, International Photographer, June 1940.
The Acme-Dunn Optical Printer
Shortly before the end of World War II, everything changed. On 18 October 1943, Linwood Dunn, first cinematographer and head of optical printing at RKO, presented to a technical conference in Hollywood “an optical printer of radically new design and construction … Besides doing all of the conventional optical printing effects, the Acme-Dunn optical printer can make automatically driven dolly or ‘zoom’ shots at any practical speed, make horizontal or vertical frame slide-off effects, wipe off in any direction at any speed, do frame-combination printing within a 12-frame cycle, and enlarge from 16mm, including successful 3-frame separation negatives16.”
The Acme-Dunn was the world’s first mass-produced optical printer. Photograph: The Cine-Technician, May/June 1944.
Built by Acme Tool Company of Burbank, the Acme-Dunn optical printer differed from the many and varied Rube Goldberg contraptions turned out by studio workshops by being constructed as a single complete unit, with a cast-iron base and housing. The Bell & Howell camera was fully integrated, and all threading and operational controls were accessible from one side of the device.
Perhaps most significantly of all, the Acme-Dunn optical printer was specifically designed for mass-production.
Describing his invention as a “dream printer”, Dunn said, “[It is] a machine which can do anything that has been done on any all-purpose optical printer, with special emphasis on ease and flexibility of operation … When an imaginative optical printing specialist is not hampered by the limitations of his equipment, his value to his studio can be tremendous17.”
All the controls for the Acme-Dunn optical printer were accessible from one side of the machine. Photograph: The Cine-Technician, May/June 1944.
Initially, the Acme-Dunn printer was manufactured purely for governmental use, with the first machine snapped up by the U.S. Navy’s Central Photographic Laboratory in Washington, D.C. After the war, widespread production began and the Acme-Dunn became what motion pictures had always lacked: an industry-standard optical printer.
On 15 March 1945, the Academy Research Council bestowed a Class 3 Award on Linwood Dunn, Cecil Love and Edward Furer for the design and construction of their new optical printer, commenting, “This machine exemplifies technical advancement necessary to keep pace with the ever increasing scope of the motion picture art18.” Nearly forty years later, in 1981, the Academy recognised the same three men for the same achievement, retrospectively awarding them a special Oscar for technical merit.
Experiments in Optical
Press advertisement from 1962 for an Acme optical printer.
Once standardised, the optical printer solidified its reputation as a piece of essential equipment capable of performing a multitude of onerous tasks without complaint — and saving the production valuable dollars to boot — as illustrated in this laconic report from a 1956 edition of Motion Picture Daily: “C&G Films Effects, New York City, announce the acquisition of a new optical printer that does everything but write dialogue … The idea, of course, is to save time in the industry where time is money19.”
Even though the optical printer was rapidly becoming an old dog, it was still capable of learning new tricks. For example, during the 1950s, Raymond Spottiswoode, an early proponent of 3D cinema, published a number of papers citing the optical printer as a useful tool in the delicate task of adjusting stereo displacement effects. And in 1957, Oxberry introduced the first commercially available aerial image optical printer, so named because the receiving camera was focused not on the plane of the film it was copying, but on a “virtual” or “aerial” image floating in empty space between its own lens and that of the projector.
Experimental filmmakers found even more creative uses for various kinds of printers. In 1963, Stan Brakhage created the short film Mothlight by pasting moth wings and other fragments from the natural world between two strips of mylar, and running the result through an contact printer20.
Watch Mothlight by Stan Brakhage:
However, even as artists were exploring the creative possibilities of the optical printer, countless machines were being mothballed as the big Hollywood studios began closing down their special effects departments. By the 1970s, the appetite of audiences for big-screen thrills had waned. Nobody cared about visual effects, and those filmmakers still passionate about them found themselves well-served by in-camera techniques such as front and rear projection, or latent image composites.
Once a highly-evolved organism, the optical printer was rapidly going the way of the dinosaur.
In 1977, Star Wars took the world by storm. To deliver the fast-moving visual effects envisioned by director George Lucas, the newly-formed Industrial Light & Magic developed a computer-controlled camera platform known as the Dykstraflex. The resulting footage was perfect for Lucas’s needs, but in order to combine the many separate elements generated by the Dykstraflex into a single image, ILM was going to need the granddaddy of all optical printers.
To create their complex composites, ILM repurposed an old VistaVision machine, originally built by Howard Anderson in the 1950s and used in the production of epics including The Ten Commandments. Resurrecting the large VistaVision format was a deliberate choice — the subsequent reduction to 35mm anamorphic in the “Anderson” optical printer helped retain the definition and clarity of the original images.
For The Empire Strikes Back, visual effects supervisor Richard Edlund championed the design and construction of a brand new aerial image optical printer. The “Quad” had no less than four projector heads, allowing many shots to be assembled in a single pass. However, the monster machine’s complexity made it difficult to load, so when ILM came to scale the visual effects mountain that was Return of the Jedi, they took a “divide and conquer” approach and split the Quad in half. One of the resulting pair of printers continued to go by the original name, while the other was christened the “Workhorse”.
Ken Smith at the controls of ILM’s “Workhorse” optical printer. Photograph courtesy of ILM.
Just as the early optical printers of the 1920s and 1930s were hand-crafted labours of love, so ILM’s machines were one-offs. Many used second-hand components, and each had its individual quirks. For example, most optical printers had the receiving camera positioned to the right of the projectors, and were thus considered “right-handed”. The Workhorse, due to the nature of the projectors scavenged for its construction, was “left-handed”; perhaps it was no coincidence that the man who oversaw its development, ILM’s optical photography supervisor Bruce Nicholson, was a southpaw too.
Having elevated optical printers into the spotlight once more, ILM continued to use them all the way up to 1993, when the venerable Anderson printer — still in working order — was finally decommissioned.
Cut and Print
Eastman-Kodak trade advertisement from 1947.
Optical printers continue to be used today by photochemical diehards, experimental filmmakers, film archivists and restorers, and in the educational market. When it comes to feature films, however, the optical printer has been almost completely supplanted by its digital successors. Yet its memory lingers in the vocabulary of the modern visual effects artist, who might casually speak of “elements”, “passes”, “mattes” and “wipes” without once reflecting on how and where the terms originated.
They might also be surprised to discover that their objective in putting together their brave new digital composites is exactly the same as that proposed by Carl Louis Gregory back in 1928, when he described the incredible accuracy necessary for the creation of convincing visual effects:
“The components must be reassembled with a mathematical precision so fine that the new combinations shall not reveal the joining lines between the welded parts even when magnified hundreds of diameters on the screen.”
Creating the joins. Not seeing the joins. Ultimately, that is the aim of all visual effects artists. And exactly what the optical printer was built to do.
Article updated July 14 2016. Special thanks to Greg Grusby and Mark Toscano. “Star Wars” photograph copyright © 1977 by Lucasfilm, Ltd.
1. Moving Picture World, 23 March 1918, “Reduction Prints for the Home Market”; 2,3. Journal of the Society of Motion Picture Engineers, May 1922; 3. Journal of the Society of Motion Picture Engineers, May 1922; 4. Journal of the Society of Motion Picture Engineers, September 1927; 5,6. Journal of the Society of Motion Picture Engineers, April 1928, “An Optical Printer for Trick Work”; 7,8. Cinematographic Annual, Volume 2, 1931, “Optical Printing”; 9. International Photographer, June 1940, “Laps ‘n Wipes”; 10. American Cinematographer, April 1934, “Tricks by Optical Printing”; 11. Journal of the Society of Motion Picture Engineers, February 1937; 12. International Photographer, December 1940, “Special Effects at RKO”; 13,14,15. International Photographer, April 1940, “Factory Methods Come to Pictures”; 16,17. Journal of the Society of Motion Picture Engineers, January 1944, “The New Acme-Dunn Optical Printer”; 18. The Film Daily, 16 March 1945; 19. Motion Picture Daily, 20 February 1956; 20. An Introduction to the American Underground Film, Sheldon Renan, first published 1967 by E.P.Dutton & Co. Inc., Released to the Internet Archive by the Author in 2011.; 21. The Making of The Empire Strikes Back, J.W. Rinzler, first published 2010 by LucasBooks.; 22. Journal of the Society of Motion Picture Engineers, April 1928, “An Optical Printer for Trick Work”