Spectral luminous efficiency based on shading perception

Purpose. We measured the spectral luminous efficiency using the criterion of the shape from shading to investigate the spectral sensitivity of the achromatic pathway at low spatial frequency. For comparison we also measured the spectral luminous efficiency of flicker photometry and brightness matching. Methods. We used a 2-channel Maxwellian-view optical system. A shading figure consisted of a reference color and a test color. Their luminance profiles were cumulative Gaussian functions along the horizontal axis with slopes in the opposite directions. When one of these colors was darker than the other, shading could be seen. The reference color was white of constant luminance and the test color was monochromatic. We varied the space constant of the function to change the spatial frequency content in the stimulus (cutoff frequency varied from 1.8 to 0.4 cpd). The observer's task was to adjust the luminance of the test color to minimize the impression of depth due to shading. We assumed that when the intensity of the test color is equal to that of the reference color for the mechanism that mediates the shading perception, shading disappears or is minimized. Results. The spectral luminous efficiency for shading disappearance was closer to that for flicker photometry than that for brightness matching. However, at short-wavelengths the sensitivity for the shading disappearance was higher than for flicker photometry and this difference increased with the increase of space constant. This higher sensitivity at short-wavelengths can be explained by the contribution of S cones. Conclusions. The results suggest that the achromatic information is mediated by luminance channel independent of spatial frequency and that the S cones contribute to the luminance channel at the low spatial frequency region.