European Conference on Visual Perception edition:33 location:Budapest date:22-26 August 2004
Brown and MacLeod (1997 Current Biology 7 844 - 849) reported that test patches embedded in a uniform neutral surround appear more saturated than the same patches presented in a variegated surround with the same average colour. We investigated how this gamut-compression effect is influenced by the purity of the test patch and the distribution of variance within the surround. In the first experiment, patches of different purities were presented in a uniform surround and matched by patches presented in a variegated surround. It was found that the gamut-expansion effect is not constant for a given surround, but decreases with increasing purity of the test patch. A further important observation, at odds with a simple explanation in terms of adaptation, was that subjects were often unable to establish a satisfactory match when the target patch was of low purity, ie when the target - surround contrast was low. In additional experiments, we varied the distribution of variance in the surround. Our results show that the influence of the surround variance is a predominantly local, and not a global, phenomenon: The gamut-compressing effect of the variance decreased rapidly with increasing retinal distance of the colour modulation from the test patch.