Abstract

From the experimental evidence, at least three processes participate in the production of the binocular response to brief changes in photopic luminance. A skeletal model is proposed, which combines three processes to form two binocular channels: (i) The fused channel—the carrier of an early binocular, excitatory process, and (ii) the either-eye channel comprised of monocular ipsilateral excitation and contralateral (reciprocal) inhibition. In a late process, (i) and (ii) together produce the common binocular output. The model accounts for the usual amount of the binocular superiority at the differential luminance threshold: a factor of 1.5. The model also allows for the vagaries of binocular superiority, which can vary from less than a factor of 1 to more than a factor of 2 in different conditions and/or in different observers. In the main experiment, we measured summation at threshold, for two-pulse luminance changes (++, −−, +−, −+) presented with a variable interpulse delay (SOA) to the same eye or to opposite keda¹ and Rashbass.² The results on interocular summation were different, but only at small SOA values. At SOA > 50 ms, thresholds became independent of stimuli polarity and port of entry. Thus, rectification (squaring?) and the final binocular combination must both occur relatively late in visual processing.

© 1985 Optical Society of America