Colorlab Preserves First Motion Image of President Taft, a Kinemacolor Treasure
Julia Nicoll, Colorlab Optical Department
with assistance from Ernest Aschenbach
 |
Left: Detail of Taft fragment. Note the difference in density between frames and the notch on the center frame. This frame is known as the "red" record (see arrows).The frame without the notch is the "green" record. These two frames "combine" to make a single frame. Right: A few frames of the Taft fragment. The entire length of the fragment supplied for preservation was 47 frames long. |
In the Fall of 2003, Colorlab was asked by the Library of Congress to do film preservation on a small Kinemacolor fragment of President William Howard Taft. The element was probably shot on October 14, 1912, according to the October 19, 1912, edition of Moving Picture World. The fragment was a 47-frame, 35mm nitrate positive that had been found by Christel Schmidt of the Library of Congress while researching the preservation of the Mary Pickford film, "In The Sultan's Garden."
Christel turned the fragment over to George Willeman, Nitrate Vault Leader at the Library's Motion Picture Conservation Center in Dayton, Ohio.. He researched and identified the Taft fragment as a Kinemacolor example. In addition to the film being labeled as the only color motion picture imagery of Taft in existence, it is one of the few remaining examples of actual Kinemacolor original positive found. Colorlab's preservation effort required duplication of the material to a stable polyester intermediate and the creation of a full-color representation of the fragment.
Kinemacolor is a curious process, evolved from the earliest efforts of film pioneers to create a "full-color" film. Before the creation of color film stocks, color could be introduced into a projection print by either toning or tinting the film. Each process had its limitations and would have to be applied to every exhibition print. Toning alters the silver remaining in the film while depositing an overall color in the film. Tinting involves using dyes to introduce color on the emulsion. The dyes could be painted on by hand or mechanically, using a stencil. A single frame could have many different areas of color. Because of the difficulty and imprecise nature of applying the tints, a tinted film does not provide the effect of what we would recognize as full-color representation. The Kinemacolor process could give the "effect" of full-color in its projection of a print by utilizing the effects of additive color synthesis.
Kinemacolor required use of a camera and projector outfitted with a special shutter that held a red and a green filter. With these filters, every other frame of the black and white panchromatic Kinemacolor camera negative film would have an alternating density. On the positive projection print, the red or green filtered light would affect the frames accordingly. The quickly alternating red and green image created in the viewer's mind the idea of full or "natural" color. Each frame of screen movement encompasses two Kinemacolor frames. In other words, the 47 frame Taft fragment really became 23 frames or nearly a second and a half of screen time when projected at 16fps.
 |
| A "green" record and a "red" record on Kodak 2383 color positive stock made from the black and white duplicate negative. The fragments represent consecutive frames and will be overlayed to make an intermediate master -- note the small differences between the frames. |
|
| Overlaying the "green" frame
wedge test over the "red" frame wedge to determine proper densities
and color rendition. |
 |
| A portion of the wedge test of red and green frames combined (in two passes) onto color positive stock before a color interpositive is created. |
Part of the preservation process for this film required the decision not to simply create a red and a green frame of the fragment but to combine the red record and green record to a single frame. The combining of the frames would allow for the fact that projectors with red and green spinning filters are not readily available, modern projectors run at standard speeds, and that color rendition in the scene we were given would appear more natural than the startling flicker given off by a red frame/green frame preservation.
In this Taft fragment, Taft appears on a ship. Sky, skin-tone, and the wooden cabin of the ship were the references that I used to judge the quality and correctness of the choices of color. A series of steps were required to fulfill the requirements of the job. After timing the positive print fragment, I created a black and white duplicate negative of the fragment on the optical printer. The liquid gate was specially adjusted to handle the 1.7% shrinkage of the element provided to us from the Library of Congress. The liquid gate also helped to create a new element with diminished appearance of scratches. The dupe neg was checked on the analyzer to make sure the timing and density of the negative were correct. I decided to select two consecutive frames to concentrate my tests on. One frame represented the "red" record; the following frame would represent the "green." Using the additive lamphouse on the optical printer, my selected "red" frame was subjected to a series of wedge tests (from 1-lite to 50) to find the right red. The same process was done to the green frame. (The red color I selected was not pure red; tests were made to find the best red that would match with the best "green," which wasn't pure green either!) The goal was ultimately to get to a color dupe negative so an intermediate stage of color interpositive was required. Kodak 2383 color positive stock was used to run tests on before going to color interpositive stage. With separate "red" and "green" wedges on color pos stock, I could overlay the consecutive frames (remember, they make up one frame), to show the variation of color density on each frame wedge and to determine which combination of red and green gave the best rendition of full color. Physically overlaying the two pieces on a light-box gave me an idea of how to start tests on the two passes each frame required. In order to make finer adjustments in color rendition, I ran a more complicated wedge test where 50 combinations of the selected red light and green light were utilized --a grid of five red light selections combined with ten green light selections. From this grid, I made my final selections. Throughout this process, the Kodak LAD process (Laboratory Aim Density) was strictly adhered to. Because the Taft fragment was so short, the sequence was shot "looped" at a 1:1 ratio at an effective speed of 16fps. Following the 1:1 duplication, the sequence of frames were stretched to a 1:2 duplication and looped. Next, we made a contact color duplicate negative from the I/P.
Then finally, we made a beautiful color 35mm color positive print of this remarkable,
92-year-old treasure from the emerging motion picture industry.
 |
 |
Left: Detail of combined and completed Taft fragment. Right: Frames of the combined Kinemacolor fragment. Combining the frames now makes the fragment 23 frames long or nearly a second and a half at 16fps. |
|
 |
Still photograph of Taft, perhaps taken the same day as the Kinemacolor fragment. Photograph provided by the Library of Congress. Title: [President William Taft on the Ship Mayflower] |
