Abstract
Introduction. The isoluminance paradigm, in which chromatically defined stimuli are matched for luminance, was developed to investigate the role of colour information in various aspects of visual processing. Performance is typically impaired, leading to the conclusion that particular tasks, e.g. motion and stereo processing, are substantially colour blind. However, other techniques that do not use isoluminance have shown that these processes are strongly sensitive to colour. One such technique is the chromatic noise-in-luminance approach: stimuli are defined by both colour and luminance, but the stimuli to be matched have opposite luminance polarities, e.g. a chromatic dot changes its luminance polarity as is moves. This leads to the question: why are some visual tasks impaired when isoluminant stimuli are used, even though the system is sensitive to colour information? Here we tested two possibilities: 1) colour sensitive cells have poor spatial resolution, and hence are impaired when fine spatial-resolution is required; 2) impairment is specific to isoluminant stimuli presented on a red background (Breitmeyer & Williams, 1990 Vis Res). Methods. Motion and form processing were assessed using global-dot-motion and Glass patterns, respectively. A temporal 2AFC procedure was used. Dot/dipole densities varied from 1 to 2 dots/deg2. The background was either achromatic (grey) or chromatic (red or green). Results. When the grey background was used, performance for the isoluminant dots was not impaired by increasing dot density. Thresholds being at least equal to the chromatic noise-in-luminance condition. Use of a red background impaired performance, particularly with the motion stimuli. Conclusions. Increasing dot density does not impair the processing of all isoluminant stimuli. However, when the stimuli and the background are both chromatic (red and green) performance is impaired, especially with motion stimuli on a red background.
Meeting abstract presented at VSS 2014