Estimation of errors in luminance signals encoded by primate retina resulting from sampling of natural images with red and green cones.

Both long-wavelength-sensitive (L) and medium-wavelength-sensitive (M) cones contribute to luminance mechanisms in human vision. This means that luminance and chromatic signals may be confounded. We use power spectra from natural images to estimate the magnitude of the corruption of luminance signals encoded by an array of retinal ganglion cells resembling the primate magnocellular neurons. The magnitude of this corruption is dependent on the cone lattice and is most severe where cones form clumps of a single spectral type. We find that chromatic corruption may equal or exceed the amplitude of other sources of noise and so could impose constraints on visual performance and on eye design.