As reported earlier, I’ve been reading Simon Ing’s The Eye: A Natural History, which is not only an exceptional, highly-readable account of the mysteries and mechanics of eyesight, but incidentally has information of use to us in the study of early film. I’ve already covered his demolition of the persistence of vision fallacy. A later chapter covers how we see colour.

Silent cinema was filled with colour. From the earliest years selected films were hand-painted, a process that was then mechanised by a system of stencils (and massed ranks of women operatives) by the Pathé and Gaumont companies. Later films were subtly tinted and toned throughout, with colours used to denote emotions as well as settings. Restorations of sophisticated colour effects in silent films are the pride and joy of film archives. There is an excellent essay by Tom Gunning, ‘Colorful Metaphors: The Attraction of Color in Early Silent Cinema’ which places early, artificial colour within the broader context of colour reproduction across other media.

However, there was also ‘natural’ colour in the silent era. The 1920s saw systems such as Prizmacolor and two-colour Technicolor in use for a handful of films, but the first of them, and the colour system with the most romantic history, was Kinemacolor. You can read all about the history of Kinemacolor, which was invented in 1906 and first exhibited in 1908, on my Charles Urban website, Urban being the entrepreneur behind Kinemacolor.

The relevance of Ings’ book here is his account of the work of Edwin H. Land. Land was an American scientist and inventor, best known for having given us the Polaroid camera. In 1959, Land devised an experiment which challenged previous theories of colour vision. This is how Ings describes the initial discovery:

On evening, at the end of a long series of experiments with three projectors [they were experimenting with red, green and blue light], Land and his assistants shut off their blue projector and took the green filter out of the green projector. Then, one of Land’s assistants, Meroe Morse, called their attention to the screen. The red projector was still running, projecting the red record on the screen in red light, and the unfiltered green projector was projecting the green record with white light. That combination of red and white lights should, in Morse’s mind, produce something pinkish. But there was the original image, its every colour still identifiable. How could red and white lights throw blues and greens on the screen?

Answer – because in colour vision, context is everything. The eye perceives colours in relation to other colours, and even though Land and his team ‘removed’ colour filters in assorted combinations, they could still generate a full colour, projected image. The eye filled in the gaps.

This property of vision, however, had been discovered and exploited fifty years earlier by Charles Urban and G.A. Smith, the inventor of Kinemacolor. Kinemacolor (patented 1906) used rotating red and green filters on both camera and projector, with panchromatic black-and-white film being used at double the conventional speed (i.e. around 32 frames per second). The result, despite some colour fringing inevitable given the separation of the red and green records, produced a record that was remarkably close to full colour. Red and green in combination covered a fair bit of the spectrum, and the eye did its best to fill in the rest. Hence eye-witnesses reported seeing blues which, according to pure Newtonian colour theory, could not be there. Ironically, Kinemacolor eventually failed as business after a 1913 court case in which its patent was ruled invalid because it claimed to show all the natural colours, but could not reproduce blue.

Today, few Kinemacolor films survive, and we can in any case only judge (or be fooled by) the colour effect if a true Kinemacolor film is projected using correct equipment. However, it is possible to approximate the effect photo-chemically or electronically, as in these frames from the surviving fragment of the famous Delhi Durbar film of 1912 (‘red’ record, ‘green’ record and composite effect):

Land knew nothing of Kinemacolor, and Ings makes no mention of it, which is a shame, because its history has much to tell us about perception, the socio-psychological construction of colour, and the wider understanding of the phenomenon of moving pictures.

If you are interested in knowing more about the history of Kinemacolor, why not download the chapter from my thesis on Urban which covers Kinemacolor, available from my personal website.

Find out more about Land’s experiments from Chris Taylor’s site, which plays with generating colour images using just red and white.

If you are heavily into optics and want to know more about Land’s work, see Gerald Huth’s Rethinking the Vision Process blog.

And, if you are in a UK university, college, school or public library, you can see sample Kinemacolor films from the BFI’s Screenonline site, together with other examples of Charles Urban’s remarkable film career.