AbstractZone theories of color vision transform R, G, B cone signals to a set of summing and differencing channels. Although the general principles of this transformation are well known, there are subtleties in the transformation which are not well understood. For example: (i) The r–g channel sends two kinds of signal, depending on the spatial frequencies of the stimulus. It sends a difference or hue signal at low spatial frequencies and an additive or achromatic signal in response to high spatial frequencies. (ii) The B‐cone contribution to the r–g channel is greater by a factor of 10 for desaturated stimuli. A quantitative saturation‐dependent mechanism can account for the B‐cone input to the r–g channel.