Abstract
Physiological measurements show two types of V1 suppression: cross-orientation and surround. Oddly, few psychophysical studies of contrast masking show any evidence of surround suppression. We measured how contrast thresholds for a Gabor target were modulated by the two types of suppression. Our results show that both types of suppression strongly elevate the detection thresholds, but have very distinct properties. For masks superimposed and coextensive with the target (overlay masking) suppression is largely independent of mask orientation and spatial frequency. But, for annular surrounds (surround masking) the suppression is sharply tuned for the same orientation (20 deg. bandwidth) and spatial frequency (1.5 octave bandwidth) as the test. We also show that, while overlay masking originates from within the target area, surround suppression originates from the extended annulus around it. For stimulus configurations that mimic the physiological arrangements, our contrast sensitivity measurements are completely consistent with the known physiology. By making use of the distinct tuning properties of the two types of suppression we showed that cross-orientation masking precedes surround masking. Our most surprising result is that while overlay masking is essentially the same at foveal and peripheral loci, surround masking is very strong in the periphery but nearly absent in the fovea. This explains the contradiction between physiology and psychophysics: the majority of psychophysical studies positioned their stimuli in the center of the visual field, while physiological recordings were mostly done for stimuli shown in the periphery. Our results indicate that surround suppression may have a different functionality than cross-orientation suppression, which is thought to be a contrast-normalization mechanism. Perhaps the iso-orientation suppression masks homogeneous peripheral regions, thereby selecting salient regions for further foveal scrutiny.