|
previous | next |
Chromaticity diagram for two display technologies
|
|
Here's a brief explanation of a chromaticity chart and the information it contains. First, a chromaticity chart is an attempt to represent colors from the 3D color space of human visual perception on a 2D graph. Obviously one of the dimensions has to go. We give up on color "intensity" and use the chart as an indication of hue and saturation. The axes of the chart can be thought of as two of the dimensions of the human visual system, denoted as (lowercase) x and y. The horseshoe-like curve is the locus of color perceptions that occur when we look at pure spectral wavelengths of light, such as from a laser. The point at the far right corresponds to light having a wavelength of 700nm (red). As the wavelength shortens, the perceptual response in x,y moves up the curve, reaches a peak at 520nm (green), then continues to the point at the bottom for 400nm light (blue). The straight line connecting the 400nm locus to the 700nm point is called the "line of purples". There is no spectral wavelength that can stimulate these colors; rather it is light that lacks wavelengths between red and blue that reside here. All of the colors that humans can perceive fall inside the spectral locus. It is an interesting feature of this chart that given two source colors, all of the colors that can be made by blending those colors in different amounts will fall on the line that connects them. An important extension of this is that the colors that can be made by blending three sources will fall inside the triangle defined by those sources. The vertices of the triangle are regarded as the primaries of that particular color system. This is very useful in predicting the colors that can be made by three different phosphors, as used in video displays. The colors inside the triangle represent the color gamut of the display, the colors that can be generated by the display. The exact primary colors for a given display technology are carefully selected to balance a set of tradeoffs between saturation, hue, and brightness. Some very successful phosphor combinations have been found over the years and are used in various broadcast television and video standards. One successful set comprises the Trinitron phosphors used in a vast number of television and computer displays. It embodies a design choice where the extent of the color gamut is diminished slightly in favor of a significantly brighter image. Because of its ubiquity, it forms the foundation for the sRGB color space, a standard used in PC and world-wide-web graphic design. The spectral locus represents the highest degree of purity possible for a color. As one moves away from this boundary toward the interior, colors become less saturated. In the center area, the colors become nearly neutral tints of gray. A display system, defined by a triangle of primaries, will select a point in the center to be the whitepoint for the display. It need not be the geometric center of the triangle, and for many systems it can be quite arbitrary, as the visual system will adapt, attempting to make an image look "natural". The exact mechanism of adaptation is complex and the subject of current color research, but enough is known that it is now possible to translate images from one display system to another, or to a hardcopy print, and retain the natural appearance. The circles in the triangles above represent the whitepoints for the two different display systems that are charted. A set of primaries and a whitepoint are enough to define a (linear) colorspace. These are usually called RGB colorspaces because the primaries for most useful systems are distinctly red, green and blue. A nice characteristic of these colorspaces is that they can be implemented using linear algrebra. Converting from one space to another is a matter of applying the correct 3x3 matrix operation. There is one more characteristic of most displays that modifies the colorspace (and destroys its linearity): gamma. But that's a topic for another time.
|
 home |
previous | next |