Scientists from Berkeley and Seattle this month announced they had described a new colour in Science Advances, in a discovery widely covered in the media.
Professor John Mollon FRS is an expert in visual perception and so we asked him to assess these claims for us. Here is what he told us:
What is described is not a new hue. Nor, probably, is it a colour that no one has ever previously seen. It is a teal, or peacock green, but one that is more saturated than any colours we typically might meet in the normal world.
To understand the technical achievement of the authors of this new paper (and it is an impressive achievement), it is important to know a little about the physiology of colour perception. Our colour vision depends on the presence in the retina of three types of light-absorbing cone cells, which have their peak sensitivities in the violet, the green and the yellow-green parts of the spectrum (see figure). They are referred to respectively as 'short-wave', 'middle-wave', and 'long-wave' cones. Notice that their three absorption curves overlap, and therefore, real-world stimuli never stimulate a single class of cone. What tells us about colour is the ratio of absorptions in different types of cones. Indeed, the retina does not transmit to the brain the absolute levels of the excitations of individual cone types. Within the retina there are higher-order cells that extract the ratios of excitation of different cones, and it is the signals of these cells that are sent onwards to the brain.
What the scientists from Berkeley and Seattle have done is produce ratios higher than those that are experienced in everyday life. They achieve this by stimulating only a chosen subset of cones in a local patch of retina, all of them of a single type. This cannot happen outside the laboratory, owing to overlap of the absorption curves of different cones and owing to the eye's constant fine motion, and the aberrations of its optics, which spread the image of even the tiniest light over many cones. So, an extraordinarily complex optical system is needed to correct the aberrations of the eye's optics, to stabilise the image on the retina, and to deliver a laser beam to selected cones within a patch. Even then, owing to residual errors, the isolation of a single class of cone is not perfect. Nevertheless, the retina's ratio-extracting cells will be driven beyond their normal range. When the laser beam excites (nominally) only middle-wave cones, then the observer sees the supersaturated teal. The authors call this colour olo, since the long-, middle- and short-wave cones are stimulated in (an approximation to) the ratios 0 : 1 : 0.
But there are probably simpler ways of experiencing a supersaturated colour. If, say, the eye views a blue-green wavelength after first pre-adapting to the complementary colour, then those ratio-extracting cells in the retina can be driven outside their normal range. It is an experimental question whether olo can be matched by antique manoeuvres of this kind.
