Abstract
When the S-cone signal is enhanced by intense long-wavelength chromatic adaptation, the S-cones can make a small, but significant contribution to luminance, as defined by heterochromatic flicker photometry or motion detection. Here, we investigate how the S-cone contribution to flicker photometry depends on changes either in the chromaticity or in the intensity of the adapting background. We use alternating luminance-equated tritanopic metamers in order to maintain S-cone isolation as the intensity of the adapting background is decreased (or when no background is present), and try to measure flicker photometric nulls between S-cone and L- or M-cone flickering lights by adjusting both the relative modulation and the relative phase of the two stimuli. Our results are surprising. First, we find that the same S-cone stimulus that makes a clear contribution to flicker photometry in the presence of a long-wavelength adapting background makes no measureable contribution to luminance when the background is removed. Second, we find that flicker-photometric nulls are possible between S-cone and M- and L-cone signals on backgrounds longer and equal in wavelength to 543 nm, but only if the backgrounds exceed a criterion radiance. Remarkably, these criterion radiances closely follow an M-cone spectral sensitivity. Thus, the S-cone luminance input is apparently silent unless the M-cones are also excited above a certain level. Our results suggest that the S-cone luminance signal is somehow gated by M-cone signals, a finding that is reminiscent of the silent chromatic surrounds suggested by Ingling and his co-workers.