In the other band/colour combination shown on the bottom portion of the image, the representation of surface features is markedly different. The wood lots show up in a brownish-orange colour and can often be identified by their irregular outline (as opposed to farm fields with their sharp rectangular shapes). The bare fields are grey-blue.

Why are we interested in different band/colour combinations? Because each one provides a better view of various features. Arguably, the details in cities are more distinct in the upper combination, as are the water features. Yet in the lower combination, there is a greater variety of information concerning vegetation types/densities. A good interpreter chooses that band/colour combination which serves the purpose at hand.

It should be remembered that colour composites are meant as an aid only to visual interpretation of remotely sensed imagery. It is the limitation of the human interpreter, namely his/her three-primary-colour visual system, that determines that a maximum of three bands be used in such composites. If, however, computer or automated analysis techniques are to be used, it is of little consequence (perhaps the time involved) to direct an algorithm to consider three or six or any number of bands. In the analysis of 'hyperspectral' imagery, consisting of hundreds of bands, we humans are outperformed by computers in being able to juggle such large, multi-dimensional tasks. Yet, visual interpretation allows analysts to bring 'context' into the analysis, often producing superior interpretations to those from computers.