When calculating the quantities of ink to be used in printing, in order to obtain a color viewed on a particular monitor, the profiles of the monitor and the printer we wish to use both come into play.

Let’s remind ourselves what these profiles are:

The profile of a monitor is a table (or data for an algorithm) which gives the colorimetric coordinates (e.g. Lab or XYZ) for the color produced by every set of three RGB coordinates.

The profile of the printer is a table or (or data for an algorithm) which gives the colorimetric coordinates (e.g. Lab or XYZ) of the color produced by each set of CMYK coordinates.

The process to find out the CMYK color which (on the printer in question) corresponds to a particular RGB color on the monitor in question, is as follows:

1. the colorimetric coordinates which correspond to the RGB color in question are found in the profile of the monitor;
2. these colorimetric coordinates are then located in the profile of the printer along with their corresponding CMYK percentages.

For example, if the profiles are those illustrated in the above tables, which CMYK ink combination must be printed in order to reproduce the color displayed on the monitor with RGB (120, 230, 75)?

1. from the monitor table one can ascertain that when the coordinates RGB (120, 230, 75) are input, the monitor in question displays the color with Lab colorimetric coordinates (83, -53, 63);
2. to find out the CMYK percentages for printing this color, the colorimetric coordinates Lab (83, -53, 63) are located in the profile table of the printer, and are found to correspond to the CMYK percentages (0, 44, 72, 0).

A color conversion always occurs between two profiles; the first is the source profile and the second is the destination profile. In the example given above, the source profile is that of the monitor, and the destination profile is that of the printer.

In the source profile, the table is always read from RGB to Lab (in general from device to colorimetric); in the destination profile, the table is read from Lab to CMYK (in general from colorimetric to device).

The conversion between two profiles is a general concept, which does not only come into play in the case of printing. Let us suppose that we have an RGB image displayed on a certain monitor, A. To obtain the same colors on a second monitor, B (which will have another profile), the numbers will need to be converted from the profile of monitor A to that of monitor B. The two profiles might be as follows:

The source profile, that is that of monitor A, is read from RGB to Lab (from device to colorimetric); the destination profile, monitor B, is read from Lab to RGB (from coorimetric to device).

The color produced on monitor A by the numbers RGB (120, 230, 75) has Lab coordinates (83, 53, 63) and is reproduced on monitor B with the numbers RGB (103, 230, 56); therefore, in order to obtain the identical color on monitor B it is necessary to alter the RGB numbers from (120, 230, 75) to (103, 230, 56).

In practice, the numerical conversion from source to destination is carried out by a piece of software known as the color engine.

The details of the transformation required are left to the color engine. The same transformation can produce different results if carried out by two different color engines.

A complete RGB table has more than 16 million rows and, for obvious reasons, the profile cannot contain them all. In this case, one of the tasks of the color engine is the interpolation of the table.

If, on the other hand, the profile does not contain a table, but only the information needed to construct it, the color engine has the job of calculating the values using that information.