Abstract
The relationship between perceived white and perceived luminosity is investigated by means of experiments with stimuli that recall the glare effect (Zavagno, 1999). Stimuli were presented on a CRT and consisted in two equally low luminance targets (a square and a cross) seen against 3 achromatic backgrounds that differed in intensity (0.3, 3.04, 30.2 cd/m2). The experiment was divided into 3 sessions, and each session had a different level of room illumination (dark, medium, bright). Subjects performed adjustment tasks throughout all experimental conditions. The task was to adjust the luminance of the square target until it was perceived as white, and the luminance of the cross until its center was perceived as luminous. By pressing two different keys, subjects could increase respectively the luminance of the square homogeneously, and the luminance of the square center of the cross homogeneously plus the inner side of its arms, determining on the last a luminance ramp. The results can be summarized as follows: i) threshold curves for self-luminosity are similar to those observed in previous experiments (Zavagno & Caputo, 2001); ii) they are always lower than the threshold curves for perceived white; iii) there is a linear relationship between thresholds and background luminance; iv) room illumination condition did not reach statistical significance, though a tendency can be seen. The results question the hypothesis that perceived white is the anchor for self-luminosity, and support the hypothesis that luminance gradients are critical in the determination of such percept.