The influence of Spatial arrangement of a scene in the stability of color appearance under different lights

Color rendering is defined as the \u201ceffect of an illuminant on the color appearance of objects by conscious or subconscious comparison with their color appearance under a reference illuminant\u201d[1]. A Color Rendering Index (CRI) aims to indicate through one or more values the range of the color variation of an object after changes in the illumination of the scene. In other words, it tries to quantify to what extent the chromatic appearance of an object is preserved under a given light source with reference to a standard illuminant. Goal of a CRI is to evaluate the quality of a light source given that a \u201cgood\u201d illuminant should keep color appearance as much unchanged as possible. With the arrival of novel LED based light sources, classic CRI proved not to be a measure in line with the human visual color sensation. As a follow up, a series of new methods and algorithms to compute alternative CRIs have been devised and tested, but no one has emerged as the definitive one. This paper aims at suggesting a potential weakness of all CRIs,: they get as input just the spectral power distribution (SPD) of the light source and implement a Color Adaptation Transform (CAT) as a simplification of the Human Vision System (HVS), ignoring its spatial mechanisms. In everyday life, perceptive characteristics of objects are strongly dependent on the context in which they are placed since we rarely see colors in isolation, as well as on the quality and SPD of light radiation[2]. Classic examples in which the spatial distribution of a scene plays an important role in the color appearance are the optical illusions. We aim to study if the spatial arrangement of a scene may influence the final appearance under different illuminants[3,4] and thus if it should consider it in its computation