Figure 2.
Localization of visual, auditory, and bimodal stimuli under various conditions for subject P.B. Data for the three types of stimuli are reported as psychometric curves (see legend), fitted with cumulative Gaussian functions (solid lines), the median and SD of which estimate, respectively, the PSE and the just-noticeable difference (or threshold). Below the abscissae, the lamp and the speaker indicate the positions of the visual and the auditory probe stimuli: the difference between the PSE and the position of the probe defines the localization bias. A, During fixation, unimodal localization was veridical (the visual and the auditory PSE coincide with the positions of the visual and the auditory probes), visual localization was more precise than auditory localization, and bimodal conflicting stimuli were localized where their visual component was actually displayed. B, For stimuli presented perisaccadically (in the interval of 25–0 ms before saccadic onset), the auditory localization performance remained unchanged, whereas visual localization became less precise and grossly biased (the test was mislocalized in the direction of the eye movement). Thus, vision and audition provided conflicting cues as to the location of bimodal perisaccadic stimuli (although physically congruent), like the bimodal conflicting stimuli during fixation. However, under these conditions, the perceived position of bimodal stimuli was intermediate between the perceived positions of their visual and auditory components. In this subject, the precision of visual localization during saccades got similar to that of auditory localization, and bimodal perisaccadic stimuli tended to be localized more precisely than either their unimodal component presented alone (the bimodal psychometric curve was steeper than the visual and the auditory curves).
