The concept and theory of low-dimensional spectral imaging has been established for some time.  Our first two articles are additions to this field of work. In recent years Prof. Roy Berns’ group at Munsell Color Science Laboratory (MCSL) at Rochester Institute of Technology built systems and methods to capture spectral images using trichromatic cameras.  In “Image-Based Spectral Reflectance Reconstruction Using the Matrix R Method” Yonghui Zhou and Roy S. Berns propose a new method for reconstruction of the spectra of each pixel in the image. These spectra are constructed from information of lower dimension than the number of measurement intervals in the spectrum. The authors transform from RGB to CIE tristimulus values by conventional means with explicit control over the linearization of the digital counts along the way.  Then they calculate the fundamental stimulus from that data by matrix R methods. By using the camera on a multi-target image checker with known reflectances, the authors are able to establish the spectra transformation between digital count and percent reflectance. By adding the metameric black of the digital count to the fundamental stimulus of the linearized digital count, the authors obtain a spectrum of higher fidelity with the measured spectrum than would have been obtained using conventional means.
    The MCSL systems were applied primarily to the documentation of cultural heritage including artist’s pigment spectral estimation where the universe of spectral properties is well defined and imaging performed under very controlled museum photographic studio environment. In the next article Eva M. Valero, Juan L. Nievas, Sérgio Nascimento, Kinjiro Amano, and David Foster present a different application of low-dimensional spectral estimation. “Recovering Spectral Data from Natural Scenes with an RGB Digital Camera” describes the use of least squares to recover the spectra of natural scenes using an RGB camera with added filters. In this simulation work of spectral imaging, the authors use a RGB camera with few broad band colored filters to capture spectral images. By introducing a “direct mapping method” they were able to reach quite good spectral accuracy in the simulations. It is interesting to see this type of spectral estimation techniques applied to more general natural scenes.
    “On a clear day you can see for ever” or so the song goes. One is likely to agree that on certain occasions, there is a special clarity about the objects in a scene or painting that it strikes us as something special. The impression of visual clarity probably results from a contrast between objects resulting from the illumination of the scene.This is an important characteristic of the light source. However, the current color rendering index does not adequately predict visual clarity. Kenjiro Hashimoto, Tadashi Yano, Masanori Shimizu, and Yoshinobu Nayatani present “A New Method for Specifying Color Rendering of Light Sources” in the next article in this issue. For those sources examined so far, the new FCI or “feeling of contrast” index that is derived from a transformation of a gamut area of a specially selected four color set, correlates well with the luminance ratio for sources which produce equal visual clarity.
    In scenes with mixed illumination, an observer can note that inversions can occur. For example, the brightness of an object color with low lightness under high illuminance level may be perceived higher than that of another object color with high lightness under low illuminance level. Similarly the colorfulness of object color with low chroma under high illuminance level is perceived higher than that of another object color with high chroma under low illuminance level. Our next article contains two color figures that should help the reader understand the differences between in concept between brightness and lightness, and also between colorfulness and chroma. In particular, the concept of colorfulness is important, but difficult to understand. Do not miss “On Color Appearance of Object Colors under Non-uniform Illumination and Its Complexity” by Yoshinobu Nayatani and Hideki Sakai.
    Our next article is on the “Prediction of Spectral Reflectance Factor Distribution of Color-Shift Paint.”  For many products we measure color at one specific geometry. However, we are all familiar with gonioapparent materials that change color with changes in the illumination or viewing angles. Standardizing groups such as ASTM International have struggled to write procedures that will successfully describe or measure for control purposes the color of many materials. For example ASTM E-2194 recommends the use of three geometries for metallic-flake pigmented materials. Materials that include interference pigments, require even more measurement geometries. Atsushi Takagi and Shinsuke Sato from Toyota Motor Corporation teamed with Gorow Baba from Murakami Color Research Laboratory to look at what they call “a jewel beetle” (a car) and to determine how many measurement points would be necessary to characterize its color. Readers will be pleased to find that they were able to reduce the measurements from over 48,000 (requiring 16 days to measure) to less than 1500 (requiring only 4 hours of measurements).
    Our next three articles all deal with color reproduction on output devices. The first article deals with the color gamut of a device. Knowing the color gamut of the output device will help the user determine whether a desired color can be printed or displayed accurately, or whether it will be necessary to do some type of mapping to simulate the color.  The color gamut information may be represented in several ways. In “Generic Device Color Gamut Description” Xinfeng Zhao describes two methods for device color gamut representation: one is a gamut surface description, and the other is for gamut volume description. Both methods are able to combine analytical and geometrical approaches in order to achieve higher accuracy. For printers, total ink coverage is an important parameter that should be considered in future work.
    In our next article brings us full circle by using the cameras such as those discussed in the first two articles, and going to displayed images on a CRT. Ali Yoonessi and Frederick A. A. Kingdom discuss the problem of displaying accurately colors from images taken by a calibrated digital camera. In “Faithful representation of colours on a CRT monitor,” they compared three approaches: displaying raw data, transforming the image via CIE common frame of reference, and an iterative approach that minimized the difference between the input and output image RGB values.  They found that the iterative approach produces the most faithful representation of the colors of the original image.
    We often talk about CRTs, LCDs, OLEDs and other displays, however, our next article gives an example of a very different type of display. In “Diffractive CIE 1931 Chromaticity Diagram” Joni Orava, Timo Jaaskelainen, Jussi Parkkinen, and Veli-Pekka Leppanen describe how a unique chromaticity diagram is generated by utilizing surface relief gratings on a plastic sample. When the sample is properly illuminated it will reflect the CIE chromaticity diagram with exact colors and a large gamut.
    Our last article looks at the group of pigments identified as verdigris. Verdigris are green or bluish-green pigments (all copper salts), which were used from Antiquity to the late 18th century. They were used both in easel painting, and for murals. José Manuel De La Roja, Margarita San Andrés, Natalia Sancho Cubino, and Sonia Santos Gómez participated in a project in which they reproduced a number of the recipes found in historic literature and characterized the products analytically and morphologically. They also studied the application of these pigments using different binders. These authors report their findings in “Variations in the Colorimetric Characteristics of Verdigris Pictorial Films Depending on the Process Used to Produce the Pigment and the Type of Binding Agent Used in Applying It.”
     In the Reviews Section, Maria Nadal discusses Diccionario akal del Color. In addition, I want to note that Ralph Pridmore has a brief Erratum to his article “Effects of luminance, wavelength and purity on the color attributes” which was published in Issue #3 of this year. Also we have an announcement about the Fogra Color Management Symposium.

Ellen Carter, Editor
Color Research & Application