American Cinematographer (1982) - Creating Visual Effects for Gone with the Wind
- magazine article: Creating Visual Effects for Gone with the Wind
- author(s): Clarence W.D. Slifer
- journal: American Cinematographer (01/Aug/1982)
- issue: volume 63, issue 8, page 788
- journal ISSN: 0002-7928
- publisher: American Society of Cinematographers
- keywords: Academy Awards, Alfred Hitchcock, Cannes, France, David O. Selznick, Dorothea Holt, Ingrid Bergman, J. McMillan Johnson, John F. Hamilton, Lee Garmes, New York City, New York, Paul Hill, Rebecca (1940), William Cameron Menzies
I lived with GONE WITH THE WIND almost 16 hours a day for about H months. Along with Jack Cosgrove, head of the Special Effects department of Selznick International, I was involved in the preparation and making of the picture from the beginning, in spite of it all it was a wonderful experience.
The Selznick International Company was a very young organization, and the making of a picture of that magnitude was almost staggering. Also, at the time that we were making GWTW we made INTERMEZZO, Ingrid Bergman's first American picture, and REBECCA, Alfred Hitchcock's first American picture.
Before telling you about GWTW let me brief you about the events leading up to the picture, in the spring of 19361 was sent on location by Technicolor to work on David O. Selznick's GARDEN OF ALLAH. It was only the second picture made with many exteriors using this Technicolor 3‑color process and the first to use many matte painting shots. Jack Cosgrove, ASC, was in charge of these shots for the picture. As I had worked with Jack many times from the old Pathe‑DeMille studio days, I was assigned to work with him. This was a new experience for both of us‑making matte shots in color with the 3‑strip camera. Due to the construction of the 3‑strip camera and its matte box we were forced to enclose the camera and the matte glass in a small tent‑like structure. This was very crude. So, for the next Selznick picture I designed a special matte box which could hold the Technicolor R. L. lens as well as three matte glasses and could attach to the Technicolor tripod head and require no tenting. I mention all this to show how, little by little, we were able to get together the means to be able to work on a picture like GWTW. it was most difficult to make a matte while looking through the camera" with the lens's photographic stop) since the matte changes with the stop. On a later picture I had a heavy line‑up finder made. It was calibrated to match all of the Technicolor cameras, and it could use all of the Technicolor lenses. Thus it was easy to make a matte, and while doing so we did not tie up a Technicolor camera.
Jack was a great man for spotting opportunities for matte shots and to convince the director that he needed them. So we returned from location with many shots. In those days nothing was duped, and we worked on only the original negative. Jack had taken over the old trick department that Dewey Wrigley, ASC, had had at Pathe Studio. The only thing we could use was a cement camera pedestal, a fixed painting easel for photography, a generator for lights, and the dark room. This being a color picture, and with the film balanced for daylight, we had to find incandescent lights and filters that would make them match the color temperature of daylight or arc. We found two old Mole‑Richardson rifles with 2000 watt bulbs and had filters made for them.
We were now In business. Jack did all the painting, and I did the photography. Each day I would bring out from Technicolor several scenes loaded in magazines and a Technicolor camera, D5 as I recall. It immediately became apparent that we needed an easel that would move the paintings up and down and sideways, because when the camera was taken off the camera pedestal at night for servicing and then remounted on the pedestal it did not always return to its original position, so the painting would be out of register for the camera. Also, if another camera and lens were used the size would be wrong. Another thing that was needed was a variable speed motor for critical balancing of the painting exposure against the original photography. The longer the film was held the more the latent image of the original exposure would change in density and color. We also needed a better way to project the small piece of negative of the original photography so that a pencil outline could be drawn of the scene upon a neutral painted glass or masonite board‑about 30 x 40 in size.
Jack laid out several shots and roughed in the painting for color based upon the colors seen in the short piece of film of the original photography that had been processed and printed by (he lab. We soon tested several scenes and sent them to the Technicolor lab. Liach scene was in a separate magazine holding three rolls of film. This was known as a group. These three undeveloped rolls of film had the blue, red, and green exposures of the original scene on which u portion had been matted out so that we could add to this scene by paintings, i.e., a night sky with stars, a building, or whatever was necessary. As each group was placed on or taken off this camera, it had to be carefully synched to keep the three films in register and also the proper footage noted to keep the films in sync for action.
In order to arrive at the proper balance for exposure and color, a small piece of the matted original film would be exposed of the painting with a temporary matte. This would be developed in the dark room with "hand test" developer that Technicolor furnished. At the head end of each scene, when we were on location, we had photographed about 4OO feet of film without action so that we could make hand and lab tests. On this 4OO foot piece of test film we would use about 18 feet for a composite test which would be sent to Technicolor for developing and printing. We then would take...