The border contour remaining when juxtaposed, heterochromatic fields were equated in luminance by the minimally distinct border (MDB) method was assessed by an increment-threshold technique. In general, the results showed that increment-threshold sensitivity, using a white test probe, did not vary across the chromatic border. A second control experiment, in which the MDB was deliberately offset, demonstrated that increment thresholds reflect achromatic luminance imbalances introduced between heterochromatic fields. This second experiment showed an increase of threshold at the border similar to the Mach-band-like elevations commonly found for fields of dissimilar homochromatic luminance. Both experiments yield results that are consistent with current theory. That is, equating heterochromatic luminance by the MDB technique balances achromatic channels somewhere in the visual system. The chromatic channels, on the other hand, remain imbalanced making the more-saturated field appear brighter. The present work suggests that the increment-threshold method is sensitive only to differences between achromatic channels, not to relative brightness differences due to chromatic-channel imbalance.
Frank Ward and Brian W. Tansley, "Increment thresholds across minimally distinct borders," J. Opt. Soc. Am. 64, 760-762 (1974)