Color induction was measured using a matching method for two spatial
patterns, each composed of double contours. In one pattern (the standard),
the contours had sharp edges to induce the Watercolor Effect (WCE); in the
other, the two contours had a spatial taper so that the overall profile
produced a sawtooth edge, or ramped stimulus. These patterns were chosen
based on our previous study demonstrating that the strength of the
chromatic WCE depends on a luminance difference between the two contours.
Low-pass chromatic mechanisms, unlike bandpass luminance mechanisms, may
be expected to be insensitive to the difference between the two spatial
profiles. The strength of the watercolor spreading was similar for the two
patterns at narrow widths of the contour possibly because of chromatic
aberration, but with wider contours, the standard stimulus produced
stronger assimilation than the ramped stimulus. This research suggests
that luminance-dependent chromatic mechanisms mediate the WCE and that
these mechanisms are sensitive to differences in the two spatial profiles
of the pattern contours only when they are wide.