We open this issue with the article, ‘‘The Development of Roof Color in Ancient China.’’ With advances in tiles making, the choice of the color of tiles used for roofs in China changed from a gray to very colorful. However, the color usage in ancient architecture was regulated by hierarchy and the theory of five colors and five essences. Those regulations and laws make clear the allowed roof colors in different zones and classes of structure. Aiping Gou and Jiangbo Wang report on research about the characters of the roof colors in different period, finding the turning points and reasons of roof color changes in crucial periods. They examined specific color samples of different dynasties tested on site, the related laws and documentations, as well as applying the theory of five colors and five essences.

    As we move through the centuries from our first article, which dealt with color starting about 4000 years ago, we will stop on the way for Zena O’Connor to examine early and more recent theories and definitions of color harmony. After examining the paradigms and assumptions embedded within these theories and the validity of harmony predictions, Dr. O’Connor proposes an updated definition of color harmony in her article ‘‘Colour Harmony Revisited.’’ She also proposes a conceptual model that represents an attempt to revise color harmony to fit with the current theoretical paradigms. Her model is contingent on factors that influence the relationship between color and aesthetic response such as individual and cultural differences as well as perceptual, contextual, and temporal factors.

    Our next article takes a new look at the Rochester Institute of Technology RIT-Dupont dataset, which was created in the 1980s primarily for the development and testing of color difference metrics. This data were published as 156 median tolerances. These tolerances were derived using probit analysis, where 50 observers judged the total color difference of 958 color-difference pairs in comparison to a near-neutral anchor pair. There has been an interest in making the complete visual data available, which was accomplished as described in this article, ‘‘RIT-Dupont Supra-Threshold Color-Tolerance Individual Color Difference Pair Dataset.’’ However, having this dataset available, there was a concern that it could be used improperly by assigning uniform uncertainty when by definition, visual stimuli different than the median tolerance have greater uncertainty. Therefore, in this article, Roy S. Berns and Bingxin Huo describe a method to assign a weight for each color-difference pair. It is recommended that these weights should be used for both performance evaluation, such as an F-test on WSTRESS values, and for formula development where weighted regression or nonlinear optimization is used.

    Talking about color differences, Kaida Xiao, M. Ronnier Luo, Changjun Li, and Guowei Hong describe psychophysical experiments to investigate the color appearance changes between the color perceived from small patches and real rooms. In ‘‘Colour Appearance of Room Colours,’’ they report that colors appear lighter and more colorful for room colors than for patches that subtended about 10° of visual field. They also state that the size change from a small patch to room size has a large effect on lightness and chroma attributes, but little effect on the hue attribute. In addition, the results show that the wall color has a considerable effect on the overall illuminant in a room, and that these effects are independent of the light source used. Then, the authors proposed a method for predicting the appearance of room colors by combining the illuminant prediction, color appearance model (CIECAM02), and a size effect correction.

    Naming colors is an important communication skill, learned as a child but used throughout life. It is useful to have a computational technique for identifying stimuli (materials or images) with the names for colors that people would commonly use and understand. In our next article, ‘‘Colour Category Foci of Munsell Colour Spectra Revealed by Two Computational Methods,’’ Elina Raisanen and Markku Hauta-Kasari examine to what extent certain computational methods provide results that match those of psychophysical research. They compare two methods, non-negative matrix factorization and self-organizing maps, to find computationally how Munsell color spectra actually correlate with their corresponding color names reported in psychophysical experiments. That the result remains consistent across the studies and different sample sets is particularly significant.

    Color has been used as a clue in many computer vision applications such as object recognition and tracking, scene understanding, image reproduction, and photography. Our brain has the ability to assign a constant color to an object although the actual stimuli vary. Actually, the color depends on the scene illumination, the content of the scene, and the characteristics of the sensor. Many color constancy algorithms of increasing complexity have been developed to help computers do this task which our brains do easily. However, the gray world algorithm based on the assumption that everything in a scene averages to a gray is still widely used. In our next article, Bing Li, De Xu, Weihua Xiong, and Songhe Feng present a novel iteration method to identify achromatic surface for illumination estimation and introduce the local grey-edge method to optimize the initial condition of the iteration so as to improve the accuracy of the proposed algorithm. In ‘‘Color Constancy Using Achromatic Surface,’’ they show how gray surfaces can be detected in an image using an illumination-independent descriptor. The color of these surfaces is then used to compute a color-corrected image. The experiment results on different image datasets show that their algorithm is effective and works as well or better than existing competitive methods.

    For our final article, we go to the textile industry, where the color of most fabrics is a result of a combination of two or more dyes. One of the issues that must be considered when developing the formulation of the color is the compatibility of the dyes. Haleh Khalili and Seyed Hossein Amirshahi present ‘‘A Novel Method for Determination of Compatibility of Dyes by Means of Principal Component Analysis.’’ In their method, the reflectance spectra of samples dyed with the classical dip-test method were converted to the traditional Kubelka-Munk absorption and scattering (K/S) data, and the dimensional properties of dataset were evaluated by the eigenvalues of each matrix. The percentage variance and residual percentage variance of the distribution of data around the first eigenvector were used for quantitative evaluation of binary mixtures. They found that the K/S data of the fully compatible mixtures scattered around one dimension and are proportional to the degree of deviation from compatibility. According to results, dyes with a percentage of variance value greater than 99.5% lead to excellent compatibility.

    We end with a short note on ‘‘Whiteness, Chromaticness, and Blackness Symmetries for CIELAB’’ and a book review. In an extension of his earlier article, ‘‘using symmetry to understand the attributes of color,’’ [Vol. 33, pp 27– 44, 2008], Lou Adams describes an interesting concept. Defining whiteness, blackness, and chromaticness in CIELAB can connect visual perception and color language to CIELAB coordinates in a simpler way. Then, we close this issue with Rod Heckaman’s review of Fundamentals of Digital Imaging by H.J. Trussell and M.J. Vrhel.