Recently, Logvinenko introduced a new object-color space defining a complete object-color atlas that is invariant to illumination.1 However, the existing implementation for calculating the new atlas's color descriptors is computationally expensive and does not work for all types of illuminants. A new algorithm is presented here that efficiently calculates the required color descriptors over large data sets and across a wide variety of illuminants. Its Matlab implementation has been made available online. The algorithm is then used to explore some features and possible applications of Logvinenko's color atlas. In particular, it is applied to images to investigate the perceptual correlates of the color descriptors; it is used to predict how images change under a change of scene illumination; and it is used to evaluate how changes in illumination and sensor sensitivities affect the mapping from the Munsell to NCS color atlases. (c) 2011 Wiley Periodicals, Inc. Col Res Appl, 20112; Published online 16 August 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/col.20680