1.3.2 Subtractive Colour Models

The colour model in Processing, and indeed in the vast majority of
computer applications, is additive: colours are generated by
additively mixing different intensities of primary colours. This is a
good model to use when the emission of light can be controlled
directly; however, there are many situations where light emission is
15
CC227 Creative Computing II Perception and Information Retrieval
fixed or not under direct control, and instead the reflection and
absorption of light by filters or pigments is used to generate colour.
Conceptually the simplest case of this is the use of filters to let light
through selectively; instead of having the starting state being ‘no
light’ and adding to it, filtering usually starts with white light, a
mixture of many light wavelengths (for example, from a filament
light bulb), and subtracts light components from it (by interposing a
coloured film, for instance).
Where in additive mixing the three primaries should correspond to
sources spanning as much of the colour space as possible (see
section 1.3.4 below for a more precise statement), the subtractive
primaries should correspond to filters removing one of the primaries
– or, in other words, transmitting a mixture of the other two. Thus a
set of three filters which respectively block red, green and blue light
act as the primary colours, transmitting respectively cyan, magenta
and yellow light.
Then, producing light of one of the additive primaries can be done
by applying two appropriate subtractive primary filters in
succession: we can produce red light, for example, by applying a
magenta filter (transmitting red and blue, but removing green
wavelengths from the white light) and then a yellow filter
(transmitting red and green but removing blue): the net effect of
the two filters is to allow through red light only. The other
combinations of two filters can be used to transmit green light and
blue light from a white light source.
Subtractive mixing is used in the process colour or CMYK model of
colour printing. A colour printer has four inks: one ink for each of
the subtractive primary colours, and a black ink (known in this
context as key). To create regions of other colours, the subtractive
primary colours are printed on top of each other, effectively
combining their subtractive effects; a cyan dot overprinted on a
yellow one produces a green dot, just as with the light filters.
A combination of the three subtractive primaries produces black;
however, there are a number of reasons why process colour includes
a separate black ink. Firstly, inks are not perfect subtractive
primaries, and consequently the black produced by their mixture is
not necessarily a strong, dark black; it can sometimes be a muddy
brown. This is particularly important when printing text, which has
the additional requirement of precise reproduction of fine details
(such as serifs): to produce those serifs from a mixture of three inks,
the registration (or alignment) of the three separate coloured images
would have to be extremely precise.
As well as these aesthetic aspects to the use of the key ink, there are
practical ones: a full mixture of three different inks on the paper can
either take too long to dry, or even cause the paper to fall apart.
Finally, using a black ink for black, and for darkening colours, will
be much cheaper than the mixture of three coloured inks.
While the key ink can be used to darken tones, it obviously cannot
lighten them. The way that light colours, and colours not producible
by an equal mixture of inks, are printed is through colour mixture
by averaging, discussed in the next section.