Color management is a complex problem. To find the solution, let us first consider the crux of the issue, leaving aside for the moment any less pivotal aspects.

The scenario is this: we are working on a Macintosh computer connected to a good quality monitor, equipped with all the programs typically used for graphics work: Photoshop, Illustrator, PageMaker, as well as InDesign by Adobe, Freehand by Macromedia and XPress by Quark.

The work in hand consists of images, drawings and text. The composition is finalised on the monitor and, after making up, is passed to a service which prepares the plates. At the printing works, the work is printed in four-color on an offset printer.

This is a simplified description, but it suffices to illustrate the problem. There are two key elements in the process: the monitor and the offset printer; let us examine their characteristics as regards color.

The monitor displays images as pixel matrices (e.g. a 17" monitor can display 1024 X 768 pixels) and each pixel is made up of three tiny points of light, invisible to the naked eye.

The color of the first point can range from black (when it is switched off) to a bright red (when lit up at maximum), passing through all the possible shades in between. The second point ranges from black to bright green and the third from black to bright blue. The three points which make up a pixel are called phosphors: the R phosphor, the G phosphor and the B phosphor.

By varying the brightness of the three phosphors, each pixel can be made to assume a series of colors ranging from black (all three phosphors switched off) to white (all three phosphors at maximum intensity). The three phosphors are very close to one another; so close that the naked eye cannot distinguish them one from another and their colors are therefore blended together. The blending occurs at the viewer’s retina, since the three phosphors are, in reality, completely separate and distinct. Such blending is known as additive mixture.

The printer produces colors by layering semi-transparent inks one over another. The four inks normally used are cyan, magenta, yellow and black (abbreviated CMYK). The range of colors that a particular printer is capable of producing (on that particular paper and with those particular inks) is obtained by varying the concentration of the inks (by means of a screen). The blending of the inks is not an additive mixture, as it does not occur at the retina. The inks are actually superimposed and the colors blend on the page; this is known as subtractive mixture.

At a certain point in the proceedings, the colors on the monitor (expressed in RGB) must be converted into the printer colors (CMYK), a process called four-color conversion.

The fact that colors are produced in different ways (additive blending of RGB on the monitor as opposed to subtractive blending of CMYK from the printer) is not a difficult problem to address. The real problem is another: the range of colours that the printer is capable of producing is not as wide as the range that the monitor can produce. In other words, there are colors which can be seen on the monitor (because the monitor can display them) but which cannot be printed (because the printer cannot achieve them).

This is the crux of the problem of digital color management; all others are variations of this problem, or are secondary problems with easier solutions.

Try this first test: There are some colors visible on the monitor (in fact, you’re looking at them now), but not all of them can be printed. If you move the mouse over the image, the colors which cannot be printed will disappear. (These tests simulate the function of an average monitor and an average printer).

Note that none of the colors of the bottom row can be printed. This is because they are too intense and too bright.

Strangely, neither black nor white can be printed, but this is not a paradox. The white seen on the screen is brighter (or "whiter") than the white of the paper, and therefore cannot be accurately printed. And the black produced by the monitor (when the pixels are switched off) cannot be obtained using printing inks.

Here is a second test, this one photographic. Again, wait until the image is fully loaded. Look at it carefully and then move the mouse over the image; the white areas contain colors which, although we can see them on the monitor, cannot be printed.

A third test: move the mouse over the image and you will see in white the colors which the monitor can produce but which cannot be printed. These are mostly various shades of green.

As we have seen, some colors (visible on a particular monitor) cannot be printed (on a particular printer). We must therefore be prepared to accept approximations to these colors.

We can focus better on the problem by stating the concept of "colors obtainable from a device", a concept covered by the term gamut.