Stresa, Italy
September 5-7, 2001

Tutorial 4

Colour in Computer Interfaces
Paul Lyons and Giovanni Moretti (Massey University, New Zealand)


We never see advertisements for software development staff that require expertise in colour interaction. But coloured screens are the primary sensory stimulus their software will produce, and poor colour choices can significantly reduce the usability of GUI applications that users stare at all day. We'll have problems if our designers don't understand colour. If you want to take chances with your software, leave colour choice to "talented" programmers.

This tutorial deals with colour theory and colour practice. It aims to demystify the task of choosing colours for computer interfaces.

The presenters will use a number of live demonstrations to illustrate a variety of psychophysical colour effects that influence users' perceptions of colours, particularly when they are used in combinations.

Tutorial Contents

In the first section of the tutorial, we deal with colour theory and the empirical evidence underlying it.

We begin by describing colour perception as a psychophysical phenomenon. This is a more complex view than the spectrum-based and primaries-based models of colour that are commonly taught by physicists and artists, but, together with ideas from those disciplines, it allows us to explain many aspects of colour that are often treated as impenetrable mysteries.

We discuss the difference between one-dimensional spectral descriptions of colour and three-dimensional, perceptually uniform colour spaces, and establish an unambiguous vocabulary for colours that is complete and canonical. We illustrate these ideas by a short demonstration involving spinning coloured disks. We discuss three-light experiments to produce the sensation of intermediate colours and introduce the concept of metamerism.

More on opponent colour theory, the eye (rods and cones, linear combinations of signals from the cones) and the choice of red, green and blue phosphors for colour CRTs.

We introduce colour spaces, including CIE, RGB and Munsell's. We discuss Munsell's concept of perceptual uniformity, Macadam's experiments, and the shape of the perceptual space. This leads to discussions of the shortcomings of the idealised 3D colour models (spheres, cylinders, inverted cones) that are commonly used in application colour pickers.

We describe why red, green and blue were chosen as the primaries for CRT tubes. Approximate perceptual uniformity of TV signals (reduced blue, to correspond to reduced human sensitivity to blue). We then run a short demonstration illustrating the perceptual non-uniformity of the conventional RGB colour space. This is related to the obvious non-uniformity of conventional colour-pickers.

This leads to a more general discussion of primaries; additive and subtractive primaries, and 3-primary and 5-primary models of colour. Then on to discussions of white points of CRT monitors and colour temperature.

Colour adaptation (relates to later discussion of monitor calibration)

The material in this section has been largely theoretical section, leavened with demonstrations and occasional practical hints. It provides a solid foundation for the second section of the tutorial, but is has a less tangible, but nevertheless beneficial effect. That is, to give non-specialists a deeper understanding of human colour perception, and to give them access to a more precise colour description vocabulary which will empower them in making decision about colour for computer interfaces.

The latter part of the tutorial translates the theoretical colour concepts described above into practical knowledge that is useful to the application developer

The programme ­ especially the first section - looks quite formalised, but the presenters introduce practical hints and demonstrations into the theoretical section, and theory underpins their recommendations about colour practice.


Each participant will receive a CD containing software components (DLLs) for colour space conversions, and for generating colour schemes based around perceptually uniform colour spaces. These components will be free for inclusion in applications.

Learning Outcomes

Participants in this tutorial will learn about

Timetable of Activities

Because practical work in this field is not naturally group-oriented, and because it would require access to colour-calibrated computer equipment, there are no practical session associated with this tutorial. The demonstrations that fulfil the same purpose are spread through the presentation, which follow this schedule.
History and colour theory: 1.5 hours
Colour practice: 1.5 hours

Who can benefit from this tutorial

The tutorial is aimed at non-specialists. It will foster a deeper understanding of colour in general, of colour in user interfaces in particular, and greater confidence in combining colours. It will cover some colour theory and vocabulary to assist them in their interface design. It will show them how to make practical use of colour theory in designing successful colour combinations.

Rather than emerge from the tutorial as talented colour designers, participants will emerge with enough understanding of colour and human perception, to make safe decisions about colour combinations in computer interfaces.

About the speakers

The presenters are both Computer Scientists with a specialisation in HCI, particularly colour selection for computer interfaces. They have previously presented industry seminars and academic papers on colour selection in New Zealand and the United States. One is an artist, and has industrial experience in interface design, including the difficulties in choosing interface colours that are both aesthetically pleasing and pragmatically useful.

Neither presenter is a primarily a colour expert. They are both users who were frustrated by the lack of information in the field, and decided to find out "how to do it." They therefore approach colour theory information with an interface implementation bias, and have filtered out much that pertains to other aspects of colour practice.

Last update: May 10, 2001