Illuminating colour: Do people see the light?

Anya Hurlbert
Newcastle University

Abstract:
We humans see colour because light interacts with particles and surfaces and then with the sensors in our eyes. But the meaning of colour comes about because we attribute colour to objects as an intrinsic property, and this conceptual operation requires unpicking the interaction between light and surfaces. Put another way, to perceive objects as having stable colours our visual systems must disentangle changes in illumination over time from changes in object properties. The yellowing of banana skin is meaningful when it comes about from ripening changes in pigment composition; less so, when it arises from changes in the light spectrum illuminating the banana. How does the human visual system distinguish temporal changes in illumination from changes in the material properties of objects? Does its ability to do so depend on the spectral properties of the illumination? Does the human visual system perceive temporal changes in illumination at all? These are questions that are closely tied to the question of how and to what extent the human visual system achieves colour constancy. Is colour constancy optimised for the types of illuminations and illumination changes under which the human visual system evolved? And can colour constancy survive the challenges posed by the varieties of artificial illumination spectra which now light people’s lives?

To address these questions, I will explore the spectral dynamics of natural illumination, and discuss evidence that visual and non-visual responses are tuned to these characteristics. New measurements of natural illumination, together with analysis of existing spectral irradiance databases reveal a characteristic pattern: rapid changes in chromaticity at the start and end of the day, when illuminance is lowest, with relative stability in between; with a slower, smoother rise and fall in illuminance, following the change in solar elevation, interrupted by unsystematic spikes due to weather-related factors. Behavioural experiments in which people are required to detect temporal changes in artificial illumination – generated by spectrally tuneable lamps in an immersive environment and mimicking natural illumination – reveal that the visibility of natural changes in illumination is generally low, and depends on the chromatic direction of change as well as the adapting chromaticity. Temporal changes away from extreme chromaticities (very warm or very cool chromaticities) towards neutral are significantly harder to detect than changes in the opposite direction, when illuminance is held constant. This effect reverses when illuminance changes at the same time. The results suggest that the human visual system has evolved mechanisms to dampen sensitivity to natural changes in illumination, maintaining perceptual stability of object colour. Concurrently, the non-visual system, fed by the melanopsin-containing intrinsically photosensitive retinal ganglion cells, is well suited to follow the slow, smooth changes of natural daylight. It is likely that the non-visual responses contribute to the feelings aroused by naturally changing illumination, even when those changes are visually undetectable.

Changes in illumination that violate the natural pattern are, on the other hand, readily visible, and difficult to distinguish from changes in object colour. Although these pose a challenge to colour constancy, they may also be exploited to enhance material properties, of artworks, for example, as well as the affective and aesthetic responses of the viewer.

Bio: Anya Hurlbert is Professor of Visual Neuroscience and Dean of Advancement at Newcastle University. She trained as a physicist (BA, Princeton University), physiologist (MA, Cambridge University), neuroscientist (PhD, Brain and Cognitive Sciences, MIT), and physician (MD, Harvard Medical School). After doing postdoctoral research as a Wellcome Trust Vision Research Fellow at Oxford University, she moved to Newcastle University, where she co-founded the Institute of Neuroscience in 2003, serving as its co-Director until 2014. Hurlbert’s research focuses on colour perception and its role in everyday visual and cognitive tasks, in normal development and ageing as well as in colour vision deficiency and developmental disorders. She is also interested in applied areas such as biomedical image processing, digital imaging and novel lighting technologies for enhancing mood, performance, and aesthetic experience. Professor Hurlbert is active in the public understanding of science, lectures widely on colour perception and art, and has devised and co-curated several science-based art exhibitions, including an interactive installation at the National Gallery, London, for its 2014 summer exhibition Making Colour. She is former Chairman of the Colour Group (GB) and Scientist Trustee of the National Gallery, and currently on the editorial board of the Journal of Vision, the Board of Directors of the Vision Sciences Society, and the Rank Prize Funds Optoelectronics Committee.

More information about Anya Hurlbert