Fig. 3.

Control experiments for subadditive spatial summation. A: slow event-related design using a full-field aperture and apertures covering either the left or right hemifield. Data are BOLD time series averaged across V1 voxels with pRF centers within 10 angular degrees from the vertical meridian. The response to the full aperture is smaller than the sum of the responses to the left and right apertures (dotted line), indicating subadditive summation. B: low-contrast experiment. A ceiling on the BOLD response might exist due to saturation of neural activity or limitations on hemodynamic mechanisms (Buxton et al. 2004). To test whether a response ceiling explains subadditive summation, we presented contrast patterns at low contrast. In all visual field maps, subadditive summation occurs at low contrast, and the response to the full aperture at low contrast is smaller than the corresponding response at high contrast (Supporting Table B). Hence, subadditivity at low contrast cannot be explained by a response ceiling. C: object experiment. To test whether subadditivity at high contrast is due to a response ceiling, we compared the response to the full aperture containing contrast patterns with the response to the same aperture containing objects. In all maps except V1 and V2, the object stimulus elicits a higher response than the contrast stimulus (Supporting Table B). This provides further evidence that subadditivity is not due to a response ceiling.