Figure 2.

(a,b) Results of experiment 2, in which observers adapted to bidirectional random dot stimuli. DAE is plotted as a function of the density of the dots moving 45° from vertical. Additional dots moving in the opposite direction were also present in the adapter and had a density of 1 dot per deg² (triangles), 7 dots per deg² (squares), or 30 dots per deg² (inverted triangles). The top DAE function (circles), in which there were no opposite-direction dots, is taken from experiment 1. The changes in the DAE density tuning function cannot be explained in terms of an opposing DAE induced by the opposite-direction dots. (c, d) When the small DAEs induced by the three opposite-direction densities are taken into account, there remains a clear difference in the tuning functions; thus demonstrating that changes in the DAE density tuning function are not solely attributable to a subtractive combination of the DAEs induced by the two directions.