Second-site adaptation in the red-green detection pathway: only elicited by low-spatial-frequency test stimuli.

The red-green (RG) detection mechanism was revealed by measuring threshold detection contours in the L and M cone contrast plane for sine-wave test gratings of 0.8-6 c deg-1 on bright adapting fields of yellow or red. The slope of the RG detection contours was unity, indicating that the L and M contrast signals contribute equally (with opposite signs) on both the yellow and the red fields: this reflects first-site, cone-selective adaptation. Second-site adaptation, which may reflect saturation at a color-opponent site, was evidenced by the RG detection contours being further out from the origin of the cone contrast plane on the red field than on the yellow field. Second-site adaptation was strong (3-fold) for low spatial frequency test gratings but greatly diminished by 6 c deg-1. The disappearance of second-site adaptation with increasing spatial frequency can be explained by spatial frequency channels. The most sensitive detectors may comprise a low spatial frequency channel which is susceptible to masking by the chromatic, spatial DC component of the red field. The 6 c deg-1 patterns may be detected by a less sensitive, higher frequency channel which is less affected by the uniform red field. The RG spatial frequency channels likely arise in the cortex, implicating a partially central site for the second-site effect.