Effects of Adaptation on Perceived Location for First-Order and Second-Order Visual Stimuli

Observers adapted to a stimulus consisting of two vertically separated antisymmetric Gaussian blobs. This was immediately followed by a 3-Gaussian-blob alignment task, whose outer two elements were spatially coincident with those of the adapting stimulus. The adapting antisymmetric stimulus resulted in a perceived misalignment of the central element of the test stimulus, and the magnitude of this perceived offset was established by the method of constant stimuli. The apparent offset increased as a power function of the adapting stimulus contrast at all test contrast levels. Perceived offset was greater for low-contrast test stimuli, although dependence upon the contrast of the adapting stimulus was less pronounced. When expressed as a function of adapting/test contrast ratio, data for all conditions collapsed together to form a single, saturating function which was well described by the formula k/[k'+(1/ratio)], where k and k' are constants. Thus, at high adapting/test ratios, the function saturated at an offset of k/k'. Adaptation effects were measured for luminance-defined first-order stimuli, and also stimuli defined by variations in texture contrast, which can be termed second-order. The effects of adaptation on perceived offset for second-order stimuli were at least as large as those for first-order, but little or no crossover adaptation occurred, ie adapting to a second-order antisymmetric stimulus produced no effect on a first-order test stimulus and vice versa. This suggests that the mechanisms involved in the localisation of first-order and second-order stimuli are independent.