Graduated profiling: Enumerating and generating perceptual colormaps for uncalibrated computer displays

Abstract

The importance of using perceptual colormaps for visualizing numerical data is well established in the fields of scientific visualization, computer graphics and color science and related areas of research. In practice however, the use of perceptual colormaps tends to be the exception rather than the role. In general it is difficult for end-users to find suitable colormaps. In addition, even when such colormaps are available, the inherent variability in color reproduction among computer displays makes it very difficult for the users to verify that these colormaps do indeed preserve their perceptual characteristics when used on different displays. Generally, verification requires display profiling (evaluating the display's color reproduction characteristics), using a colorimeter or a similar type of measuring device. With the growth of the Internet, and the resulting proliferation of remote, client-based displays, the profiling problem has become even more difficult, and in many cases, impossible. We present a method for enumerating and generating perceptual colormaps in such a way that ensures that the perceptual characteristics of the colormaps are maintained for over a wide range of different displays. This method constructs colormaps that are guaranteed to be "perceptually correct" for a given display by using whatever partial profile information of the display is available. We use the term "graduated profiling" describe this method of partial profiling.