Abstract
We have previously shown that the integration of non-local auditory and visual motion can best be explained by a probability summation model that uses direction-independent signals from modality specific motion detectors [Wuerger et al. (2002), Journal of Vision, 2(7), 663a]. This finding contradicts other perceptual and physiological data which suggests that auditory and visual signals are integrated at an early stage. We hypothesize that this low-level integration crucially depends on the auditory and visual signals to be co-incident and co-localized.
To test this hypothesis we measured motion detection thresholds for auditory, visual and bi-modal motion stimuli, which were presented along a horizontal arc, containing 31 LEDs and loudspeakers, spaced 5 degrees apart. Each motion signal described an arc of 90 degrees in either the left or right hemi-field in front of the observer. The auditory and visual components used for the bimodal stimuli moved independently, hence they could be in the same or different hemi-fields and move in the same or opposite direction. Thresholds were significantly reduced only when the auditory and visual motion signals move in the same direction and are located in the same hemi-field. A neural summation model explains the data for this congruent condition. In all other conditions the bi-modal thresholds could be explained by an independent-decision model.
We conclude that auditory and visual signals have to be co-localized to be integrated effectively. This is consistent with the idea that local auditory and visual signals are integrated before motion is extracted.