• In order to see colour there must be an object whose surface is coloured, light to see it by and an eye to see it with.
• A surface that is coloured may also be textured and the texture may interfere with our perception of the colour.
• The light under which the object is viewed may itself be coloured. To take daylight as an example, the time of day determines significantly how the light is coloured. Think of midday compared to sunset.
• Our eyes differ between individuals in terms of relative sensitivity and therefore how a colour is perceived.
• The peripherals used - cameras, scanners, monitor screens, proofers and printers - analyse or synthesise colour in different ways.

When a scene is reproduced, as a painting for example, one of the qualitative factors is the quality of light imparted. A mundane scene can appear quite magical when the lighting is unusual or modified by climatic conditions. This need not be the result of an artist's imagination; a photograph can have the same pictorial quality. Indeed many pictures are only notable because of the light they convey.

When we are asked to reproduce such a picture we must ensure that the reproduction of the colours is accurate otherwise we may lose the very quality that made the picture worth reproducing. When reproducing pictures for advertising, the client is entitled to expect that the result will convey an accurate impression of the advertised object.

Colour printing has benefited hugely from standardisation, thus limiting some of the variables. Standard colour sets (of inks) have been adopted which enable print buyers to transfer jobs between printers with some confidence. Standard print targets and swatches printed along with each job enable printers to monitor their print characteristic and to show that the print run conforms to a standard printing characteristic. Throughout the run the targets will also ensure that the press has remained stable so that sheet-to-sheet consistency is maintained. In the past it was not unusual for an important print buyer to preside over the print run and insist that colours be printed 'heavier' or 'lighter' to achieve their idea of how the job should appear. Such changes destroy the standardisation of the print characteristic.

The International Colour Consortium (ICC) has introduced colour management, a method by which many of the problems of accurate colour reproduction have been addressed. Colour management looks after the inter-device colour accuracy, but if the appraisal of colour is not undertaken in a disciplined way, the benefits of colour management will be compromised.

Colour matching

Many industries use colour matching as part of their manufacturing process. Almost anything that is coloured and (hu)man-made is subject to colour matching checks during manufacture. The most obvious industries are paint and ink, plastics, and paper. Less obvious are the food industries. Baking and brewing make extensive use of colour matching as part of their baking and quality control processes. These industries have found colour to be an important part of their offering to the consumer, and an aid to the efficacy of their product. Colour is often specified in quality control procedures.

Colour matching has traditionally been carried out by experienced personnel comparing the colour of a product against a reference colour by eye. In the colour printing industry, colour matching the proof or the final printed sheet to the original has always been the aim.

For special colours, the PANTONE matching system has been used by designers to convey the required colour through all stages of colour reproduction with the printer mixing suitable ink colours to match the specified colour. Whilst printers are expected to weigh out pigments in order to achieve the Pantone colours, in practice most will mix a special batch of ink for the job. The mixing is undertaken by the printer and the mix is usually 'dabbed out' and visually compared to the Pantone swatch. Printers mostly involved with packaging often use a computerised 'ink kitchen' where the required colour is measured and the computer works out how much of each of a stock of ink colours should be mixed to attain the target colour.

If the job is constrained to the standard (four) process colours, then a prepress worker will estimate the dot values of yellow, magenta, cyan and black halftone to be used. Actually it is never necessary to use all four colours and the mix will consist of three colours at most.

When we reproduce a colour original (transparency, photograph or artwork), we ensure that the print 'matches' the original in terms of colour. In all cases the colour matching is checked by way of a visual comparison between the original and the print. When the print or a proof is submitted for approval, the client decides the acceptability of the submission by eye.

In printing we know in advance of starting many jobs that accurate colour matching cannot be provided. Some colours, either in the original or as specified by a Pantone reference, cannot be achieved by the four-colour process. If it is considered important enough, an extra colour can be printed so that photo-originals can be produced by the standard four colours and the fifth colour mixed to match a brand colour perhaps not attainable with four-colour process. The alternative is to make the closest colour match possible if that is acceptable. I have often thought that colour matching a photo-original is less important than producing a 'good-looking' picture and many times at the viewing booth the client only refers to the original if the printed picture looks poor.

The operating systems of computers used in the graphic arts are capable of supporting the core of the ICC colour management system, the Colour Management Module (CMM) and the Profile Connection Space (PCS). By attaching Profiles to the colour data file at particular points in the repro chain, the colour characteristics of the various hardware devices can be used to modify the colour data in order to manifest the image accurately.

When a transparency is scanned, the resultant file is dependent on the transparency material characteristic and the colour reproduction accuracy of the scanner. If these can be measured then the data file can be modified to bring the data back to what it would have been if both the material and the scanner were 'neutral' in terms of colour reproduction.

The image file will usually be displayed on a colour monitor and the characteristics of the screen should be optimised. If any of the colours in the image are modified to enhance the image, the judgement of the changes will be made on the screen and it is important that the characteristics of the screen are taken into account. If the file is sent for approval, the recipient will probably display the image on a different type of screen and this too should be profiled.

Light

We generally regard the light we see by as being white. Light has no colour itself (we cannot see it in transit) and its presence is only made apparent when it is reflected or radiated from a surface. Since our eyes adapt to the ambient light, we regard it as white and that enables us to perceive colours by reference to the white. That we reference white in this way means that we can readily determine the accuracy of greys. Grey is dark white and we can detect any colour difference between the white and grey if the grey is tinted - warm or cool greys, for example.

The colour of light has a profound effect on our perception of colours viewed. Under yellowish tungsten light we tend to confuse yellow colours with white, or dark blue colours appear to be black, for instance.

When we compare two coloured objects it is essential that we view both colours under the same light. We cannot hold a colour sample up to the monitor screen in order to compare their colours as the sample is not illuminated by the screen and the screen is self-illuminated.

When comparing two colours we hold them in close proximity and examine them for similarity. This is a valid thing to do. Any difference in the colours will be discernible by most observers and the more practised observers will be able to describe the differences in a meaningful way.

Most of us work in an environment of mixed lighting conditions - perhaps in an office with daylight coming from a window, fluorescent general lighting and a desk lamp with a tungsten lamp. Add to this the computer screen and it can be appreciated that the eye is constantly re-adapting itself to changes in light - both level and colour.

Standard lighting conditions specified in ISO 3664 (BS 950 part 2 is similar) specifies the intensity and colour of lighting for the purpose of colour appraisal in the printing industry. Viewing booths are available which conform to these specifications and are recommended where colour accuracy is important. A booth shades the viewing area from ambient light and contains suitable lighting to meet the needs of the standard. Comparisons of original photographs, proof and printed sheet should be made in the booth and transparencies can be viewed with the addition of a lightbox that conforms to the same specification.

There is a difficulty with monitors in that the phosphors of the screen are red, green and blue (check a white area of the screen using a magnifying glass) so that the colour temperature of the screen can only be approximated. The intensity of most monitors cannot approach the specified light level and the higher the screen resolution the lower the intensity as a general rule. For this reason, if a colour comparison is to be made between the monitor and a photograph, proof or print, a separate viewing booth is necessary in which the intensity can be lowered to match the monitor. Dimmable booths are available where the act of dimming the light intensity does not change the colour temperature of the illuminant.

Where colour measurement is used it is important to remember that colour is a perceived entity and no numeric representation imparts any meaningful notion of what the colour is. It is what we actually see that is important.

When we need to purchase something of a particular colour, most of us are astute enough to carry a reference colour in order to directly compare with the intended object colour. When comparing our reference and the proposed purchase we tend to move both around so as to 'catch' the light on their surfaces in order to make the best comparison. I have seen a woman in a clothes shop carry the intended purchase towards the glass door to use daylight, rather than the fluorescent lighting of the store, to make the comparison. When making such exacting use of light we instinctively 'know' whether the light is suitable or not.

Controlled lighting conditions are imperative when comparing and assessing colour.

The eye

The function of the eye is to detect light - those wavelengths of the electro-magnetic spectrum that lie between 380-700 nanometres and we call visible. The sensors in the eye are of two types: the rods, which are most sensitive to low light conditions and the cones, which are sensitive to higher light levels and are sensitive to three groups of wavelengths of light - red, green and blue. These sensors are arranged on the retina in the back of the eye facing the lens (the cornea) in a camera-like structure. As cameras go, the eye is not of especially high quality. It is, however, enhanced by the visual cortex, which performs many higher-level functions including image processing and pattern recognition. The sensitivity of the rods peaks at about 555 nanometres (blue-green) and if we consider moonlight there is a kind of blue-green eeriness associated with it. Film-makers use a blue-green filter to taint their night shots in order to convey the feeling of darkness, and we know without being told that this is night-time.

The cones are sensitive, as we said, to red, green and blue light and they are responsible for colour vision. In daylight we are able to see colours in the landscape with great discrimination. There are small differences in the sensitivities of these sensors between individuals so we see colours slightly differently, although the majority of us conform to the Standard Observer defined by the Commision Internationale d'Eclairage (CIE) in 1931.

Defects in colour vision, sufficient to cause problems, are present in about 10 per cent of the male and 1 per cent of the female population. Tests for colour vision defects are carried out in school populations where children are asked to recognise a number formed by coloured circles in a background of differently coloured circles. In these tests, developed by Ishihara for transport workers, the recognition of four correct targets is sufficient to be 90 per cent certain that defects are not present. I wonder how many people working with colour have been tested for colour vision. Most colour vision defects are inherited but chronic disease, such as diabetes, can cause a degeneration of colour vision over time.

Apart from colour perception the cones give us acuity - high levels of detail in what we see. The cones are most abundant in an area called the fovea, slightly off centre to the optical axis of the eye, and primarily responsible for what we 'see'. Although our field of vision covers about 170 degrees in span, it is only the very centre of this that affords us high resolution. We have to scan a scene to take it all in. You have to scan these lines of text to read them. If you fix your gaze on one word you can see that the surrounding words are less distinct.

The eye is particularly susceptible to fatigue and staring at a coloured object for a short time results in an after-image of complementary colour in the field of view. For this reason, the brain demands that the eye is constantly moved to present a new image to the retina. Try staring at a fixed object and within a very short time the image appears to move. Actually the brain is responsible for involuntary eye movements in order to change the image.

Our eyes are constantly adapting to the changes in light level and quality. If we move from an area of high intensity to a lower light level (go indoors on a sunny day), it takes a perceptible time for our eyes to accustom themselves to the new level. Also, if we move from an area of tungsten lighting to fluorescent lighting we often perceive the difference in colour. Move back and the yellowness of the tungsten lighting is quite apparent.

If we look at a colour monitor screen when it is off, it is grey. When we turn the screen on we are adding light to the screen. How then do we see black as part of the image? Our eyes have adapted to the white of the screen and the grey screen is dark enough to be perceived as black. Look up from the screen and the surrounding area is unchanged.

Our perception of a colour is severely affected by the colour of its immediate surroundings. This can be used to good effect in order to accentuate a colour but poor (design) use of adjacent colours can cause problems to viewers.

Implementation

If we put all this together, we can determine suitable rules for colour appraisal and matching which, if followed, optimise the chance of success. When attempting to match the colour of a sample or the reproduction of a photograph:

• Use a purpose-designed area shaded from ambient light and illuminated by light conforming to ISO 3664.
• The immediate background should be neutral in colour (grey) and should reflect 20 per cent of the light falling on it.
• The ambient light should be less bright than the illuminated area.
•  Reflections from the surface of any viewed proof, sample or original should be minimised by suitable angling of the light source and the viewing plane.
• Be wary of comparing colours when they obviously differ in gloss.
• The eye should not be fixed on a particular point for any length of time, especially if saturated colours are part of the appraisal.

When assessing colour on a monitor screen:

• The white of the monitor must be the brightest white in the field of vision. In practice this means working in subdued ambient light.
• The surrounding area should be neutral in colour.
• Daylight should be excluded.
• Do not compare the colour on the monitor with any external colour unless a specially designed viewing cabinet is used.
• Check when the monitor is off that no extraneous light falls on the surface of the screen.
• Ensure that the monitor is calibrated and that the calibration is valid.
• Ensure that the correct colour profile is used for the screen.
• When seated at the monitor, look beyond it and check that no other light source is in the field of vision.