Today we’ll explore Hollywood’s history of visual trickery with the backstory of modern greenscreen compositing.
With cameras and computers everywhere in our modern world, it’s easy to forget that the very first motion pictures were, themselves, essentially a special effect. It’s here at the beginning of filmmaking that we’ll start our journey: the close of the 19th century with one of the world’s first prolific filmmakers – a man who spent his life studying the art of illusion – Georges Méliès.
In his 1898 film “Four Haads are better than one”, Méliès employs a visual trick that is the rudimentary beginnings of what we now think of as greenscreen compositing.
The use of mattes for multiple exposures.
Compositing is a technique combining different shots and elements into one image. The matte shot was the first compositing techniques employed by early filmmakers such as Melies.
In his film, Melies would black out parts of the frame using a piece of glass with some black paint. This “matte” made it so no light would reach the film so it wouldn’t get exposed. Then Melies would rewind the film and this time matted out of everything else and expose only the part of the frame that was under the matte earlier. The resulting double exposure could combine two or more different shots into one frame all done inside the camera.
This matte technique was used again on Edwin S. Porter’s 1903 The Great Train Robbery but this time not as magic trick but as a means to create a larger more realistic world
Notice the train moving outside the window of the train station – also the open door of the mail car with the scenery in the background. Both of these shots were done using mattes and double exposure.
Now the fair question to ask here is why didn’t they just shoot it in a real train station or a real train car?
The answer is it was technically impossible at the time. Early orthochromatic film needed a lot of light and the technology for efficient electrical lighting for film was still a decade or two away. That’s not even considering the inherent exposure problems of shooting an interior scene with a window in the shot. Even modern day cameras have trouble with the brightness differences between interiors and exteriors. In order to make film behave they way we experience the world, visual trickery had to be done. As film grew up in the 1900s and 1910s more techniques for augmenting sets and creating false realities would be developed. The Glass shot was a technique of painting elements on a piece of glass and placing that glass between the subject and the camera – a sort of real world compositing which was refined by early filmmaker Norman Dawn, using it to augment sets making them look much bigger and more elaborate without the costs of construction.
But the problem with the glass shot was the paintings had to be ready on set. Norman Dawn solved this problem by painting the glass black and treating the shot like a matte shot.
The matted film would be transfer to a second camera where matte artists could take their time creating the matte paintings. This matte painting concept continued seeing use in the golden era of Hollywood and continues with us even in our digital world.
The problem with mattes is the camera had to stay perfectly still and no action could cross the matte line – the “hopefully” invisible line between the live action and the matte painting.
This is where the traveling matte came into place. The process patented by Frank Williams in 1918 and demonstrated here in F.W. Murnau’s 1927 film”Sunrise” – was a black matting process which photographed subjects against a pure black background. The film would then be copied to increasingly high contrast negatives until a black and white silhouette emerged.
This black and white silhouette was used as the matte – called a traveling matte because it moved throughout the frame.
This “black back matte” effect which was called the Williams Process was used quite famously by John P. Fulton in 1933 for the film “The Invisble Man”. The shots where the invisible man was taking off his clothes were accomplished by photographing actor Claude Rains wearing a full black velvet suit standing against a black background. This effect was so memorable and startling it was used on follow up sequels even after more effective processes came along.
The Williams Process had some issues – for one, any shadows on the subject would be lost in the traveling matte. An alternative came about in 1925, invented C. Dodge Dunning which would eventually be called the Dunning Process. This process used colored lights, lighting a background screen blue and the foreground subject in yellow. Using dyes and filters, the blue and yellow light could be split apart to create traveling mattes.
The Dunning Process would first see use on King Kong in 1933.
The problem with the Dunning Process was it only worked with black and white film. Color Film needed a new technique and it would come in 1940 by special effects artist Larry Butler in the Thief of Bagdad.
Using the three strip technicolor process, Butler shot the subject against a blue background.
Blue was used because it was the farthest away from skin tones and the blue film stock had the smallest grain. Taking the blue separation from the three technicolor negatives, Butler was able to create a silhouette matte just like with Williams process. Then, using an optical printer, a relatively new invention at the time that could combine multiple film strips into one, Butler would first remove the blue background from the foreground plate and, using the negative of the travelling matte, remove the foreground space from the background plate and then finally combining both foreground and background plates together.
This bluescreen technique won an Academy award for Best Special Effects for Larry Butler in 1940 but it was not without its inherent problems. Firstly the process was extremely time consuming as it involved several steps with an optical printer. Secondly, it still had some edge issues where a thin blue line was almost always visible in the shots. It also couldn’t handle any fine details like hair or smoke or motion blur. Despite these limitations, the blue screen process was used extensively including in such blockbusters as The Ten Commandments in 1956:
Hollywood kept experimenting with other variations on the bluescreen process including the ultraviolet matte as used in “The Old Man and The Sea”. But the real challenger to blue screen was created in the late 50s and credited to one of the giants in world of compositing Petro Vlahos.
Developed by Vlahos in the mid 50s and used extensively by the Walt Disney Studios in the 60s and 70s: The Sodium vapor process used actors, who were lit normally, standing in front of a white screen which was lit by powerful sodium vapor lights – those are the orange lights you see on street corners. Sodium vapor emits light in a very specific wavelength – averaging 589.3 nanometers – and nothing else.
Using a specially coated prism in an old three strip Technicolor camera, the very specific wavelength of the sodium vapor light was split off and captured on special black and white film – automatically creating the black and white traveling matte. The remaining light would be captured by regular three strip Technicolor Film which was relatively unaffected by the yellow/orange sodium vapor lights.
This technique produced some of the best travelling mattes of the time and was used by Disney first on film The Parent Trap and then The Absent Minded Professor both in 1961. Mary Poppins in 1964 demonstrated the capability of the sodium vapor process winning an academy award for best special effects.
There was just one problem. Only One Sodium Vapor prism was ever made so there was only one camera that was capable of this process. Disney owned the camera and they didn’t let it rent for cheap. Revenge of the Blue Screen
In the late 50s, When MGM was ready to produce Ben Hur in the MGM Camera 65 format (a 65mm film process) they turned to Petro Vlahos, the inventor of the sodium vapor process for help on the compositing. They didn’t want the problems that Ten Commandments had with bluescreen but The sodium vapor process wouldn’t work as it prism it used was been made for 35mm film, not 65mm. So Vlahos was asked to see if he could do something about trying to improve the bluescreen process.
After six months of hard work, Vlahos had a discovery. And this is where it gets pretty complicated. Most colors that aren’t purely green or purely blue have about equal amounts of blue and green in them. So when creating a matte from bluescreen, Vlahos used a Green Cancellation separation (or positive), ran it though with the original color negative exposing both pieces of film together under a blue to light to create a “blue difference matte”. This matte was clear where the blue and green were the same – Then the blue separation positive was combined with the original negative and exposed under red light to get a cover matte. This cover matte was applied back to the original color separations except that the blue separation was replaced with a composite of the green and the green difference mask – essentially a synthetic blue separation.
This complicated process required 12 film elements to get from the composite negative to the composite internegative but it was remarkable in the way it single handedly solved the edge and fine detail problems that plagued blue screen.
It was so successful in fact that the process remained in popular use for almost 40 years.
Developments like microprocessor controlled quad optical printers, employed by Richard Edlund for The Empire Strikes Back made the process faster and more accurate but the next big change to come would be in the form of digital.
I have consciously avoided the the term “chroma key” as historically the term applied only to video systems only. That’s not the case anymore. In rudimentary video mixers, a keyer was a mathematical process that would make a range of colors in a video signal and make it transparent. This is, of course, a common effect that television newsrooms all over would use weather map special effects.
Blue as a screen color was still predominate but green started to take over as films began getting the digital post production treatment in the late nineties. Why Green? Basically Green was easier and cheaper to light than blue, green registers brighter on electronic displays, worked well for outdoor keys (where the blue screen might match the sky) and the bold green color was less common in costumes than blue is.
And now as digital camera are replacing film, many digital sensors use a Bayer Pattern which have twice the number of green photosites than red or blue to capture luminance. This makes modern digital cameras much more sensitive to the green part of the spectrum making pulling a matte from greenscreen a little easier. Blue is still commonly used as are other colors depending on the needs of the shot.
So now with advanced software and motion controlled cameras, Chroma Key, a term that has grown now to encapsulate much more than it’s original video technique, can be used to insert backgrounds and set extensions in ways that Georges Milies and Norman Dawn could only dream of.
There are cynics today that believe modern film is too reliant on CGI and that we should return to a simpler form of real filmmaking. But as I hope you learned, that era never existed – filmmakers from the very beginning have sought to push the medium with special effects.
The undeniable truth about filmmaking is the only thing that matters is what’s on that screen. From Edwin S. Porter’s matted train station window to the modern action spectacle, it’s all about creating a window onto another world. A world where each of us can find our dreams our fears and ourselves. All these effects we have are just tools to help us get there.. And we have some fantastic tools, so use them, and make something great.