Copying artwork scientifically doesn't just involve the two "images" of original and copy. The reproduction process involves image versions in pigment in the original art, in phosphers in a computer monitor screen, in dyes in photographic film, in numerical values in digital files in different colour space definitions, and finally in the inks and pigments that comprise the "reproduction".

There is also another fundamental and relevant image, namely the one we generate mentally, the one we "see". It can be a hard fact to accept, but nothing is actually "coloured". Colour does not exist in objects, we generate the sensation of colour psychologically in our eye/brain system. It's true that objects reflect wavelengths received from light sources to our eyes. These wavelengths however are colourless, they do not possess a "colour" property. One of our great scientists, Sir Isaac Newton, pointed this out in the 16th Century - "... the raye itself is not colored".

So art reproduction is really about producing a copy that triggers a similar response in an observer's brain to the one they would have in looking at the original. Quantifying a psychological perception sounds a daunting task, but colour scientists have actually done pretty well in defining colour models for us to work with that compensate for the way human colour perception works.

In copying artwork, it's important to understand the balance of wavelengths in the light source illuminating the art, how the different pigments in the art absorb and reflect the wavelengths, and the sensitivity to these reflected wavelengths of the recording medium, whether it's photographic film or a digital CCD (the light sensitive part of a digital camera or scanner). And also of course, being aware of the fact that our visual reaction to these reflected wavelengths will be different. We don't just generate the same colour based on the same wavelength, the colour we "see" depends largely on surrounding colour and brightness.

Given all these problems, and the need to know what is actually happening, the way forward is to be able to measure colour scientifically with an instrument. This is a spectrophotometer. In the same way that a photographic light meter measures light brightness and gives a result calibrated against a known standard, a spectrophotometer measures small bands (typically thirty) of visible wavelengths. A colorimeter does a similar job, but just using three bands. So it is Colorimetry (colour science) that gives the control needed to produce imperceptible copies, compared to the traditional use of photographic science (Densitometry) in which the colour values are not separated from brightness.

Measurement enables the creation of files that adapt the varying colour responses of cameras, monitors, and printers to return the image to a known standard. With an instrument we can compare an original colour with its reproduction, and get units of colour difference on a scale called deltaE. Up to 2 units are considered imperceptible to human colour vision. From 2 to 5/6 is considered acceptable, and more than this increasingly unacceptable. In the colour science lab, accuracies of less than 1 unit have been achieved. Commercially, accuracy varies considerably. Mass reproductions in magazines and books are limited by production costs and use cheaper inks and papers that have relatively limited gamut (range of possible colours).

The new kid on the block is digital inkjet printing. Because it is not designed for mass print runs, much better quality inks, pigments, papers and paper receiver coatings can be used.

Tony Riley

Last revised Dec 2006