Early color cinematography process (1908) — additive color mixing via color screens in camera and projector. Alternative to Kinemacolor, technically complex but less commercially viable.
Around 1908, French engineers—primarily Édouard Berthon and Daniel Keller-Dorian—attempted to crack the problem of color recording. Their approach was based on additive color mixing: instead of chemically burning colors into the emulsion, they placed a fine color grid directly onto the film material. The grid consisted of tiny red, green, and blue lines or dots—similar to the Kinemacolor principle, but with a crucial distinction in technical implementation.
In practice, the process worked as follows: during recording, the grid was positioned between the lens and the film. Each pixel registered only the amount of light passing through its respective grid area. During projection, the exact same grid had to be positioned in front of the projection light—synchronized with the moving film. Incorrect positioning or synchronization problems immediately destroyed the entire color effect. This was the Achilles' heel: while Kinemacolor worked with two colors and was therefore more tolerant, Berthon/Keller-Dorian demanded constant mechanical precision throughout the entire projector-film chain. On modern sets, one would say: zero-margin tolerance.
Why didn't the process catch on? The answer lies in cost and reliability. Cinema owners had to equip their projectors with special grids. The film itself was expensive—the grid required precise manufacturing. And in practice: a dirty or shifted grid plate led to color flickering or color casts that disturbed audiences. Kinemacolor was simpler, more robust, cheaper—which is why it dominated the market for additive color processes in the 1910s, until subtractive processes like Technicolor later took over.
From a camera technical perspective, it's interesting: the Berthon/Keller-Dorian system forces you to think like the projector even during setup. The grid position was not cut-variable but a hardware constant. There was no flexibility like with later multi-layer processes. Those who worked with it—primarily French and British studios in the early 1910s—had to think through each shot twice: image composition AND grid compatibility. This was a conceptual precursor to later format-specific ways of thinking, such as VistaVision or Cinemascope.