Brightness contrast induced red/green balance shifts in real-world objects

In order to test if red/green balance shifts induced by bright and dark surrounds persist under real-world viewing conditions, we had 12 subjects set red/green balance matches to real-world objects (20×20×7.5cm boxes) that were brown, orange, sky blue, or blue. These boxes were presented in a chamber with either black or white surfaces, which was illuminated from above by a bright D65 daylight-simulating light. The boxes could also be covered with either white or black paper with a small 3×3cm cutout that allowed subjects to view only the center of the top surface of the colored box. Subjects set multiple matches to each condition on an adjacent LCD monitor over the course of multiple days. The results indicate that a significant induction effect occurs only in the conditions where the boxes are covered with white or black paper with the top-central cutout, and only for those boxes colored brown, orange, or sky blue (in the opposite direction of brown and orange). This is partially in line with past research showing that these effects are weak for distal surrounds and objects with relatively high S-cone activity (blue objects); however, the direction of the effects shown in this experiment is the opposite of that seen in previous experiments, and the sky blue box also shows a significant opposite direction effect. This suggests that under at least some viewing conditions in the real world, these induction effects can occur, but they are not necessarily equivalent to those seen in monitor-based experiments.