Traditional methods for colour quantification are complicated by the fact that colours change depending on illumination, and that different observers often perceive colours differently. Here we describe a new affordable method, which improves methods relying on human observers, to quantify patterns and colour variations. The procedure combines customized software with the use of digital cameras and commercial photofinishing software. The computer routines correct unavoidable illumination changes during image capturing, making all images comparable. Colours are quantified in a continuous scale of the conventional colour models developed for the human vision system, such as HSB, RGB, CMYK, or Lab, amenable for statistical analyses. We illustrate the use of this technique showing a previously unknown sexual dimorphism in the red-legged partridge, Alectoris rufa, undetectable with the unaided human eye. We also demonstrate that the digital system provides a finer discrimination than human observers for scoring the plumage of partridges belonging to two different subspecies. This method has potential applications in behavioural ecology, physiology, genetics, evolutionary biology, and taxonomy.