Digital Color Model Puts Color in a New Light by Ken Davies
The COLORCUBE is a three-dimensional model by which one can understand and teach digital color theory. This elegant representation of color bridges the gap between additive and subtractive systems of color, and defines the method by which colors are stored, manipulated, and reproduced using computer technology.
The COLORCUBE is patented in the United States and patent-pending internationally.
Return to Articles Menu
More consumers than ever are buying their way into the digital imaging market. Digital cameras, color printers and color scanners have become less expensive and therefore, more accessible to new users. Accompanying this revolution in color usage is the need to understand digital color and its inherent complexity.
Research indicates that typical end-users are baffled by the intricate behavior of color and often complain that "the colors that print do not match what is on the monitor".
In spite of astounding technological advances in color, it is readily apparent that few people understand the theory of how digital color works. This inability to fully comprehend new color technologies can lead to customer dissatisfaction and products that fall short of user expectations.
Spittin' Image Software introduces a new "low-tech" invention designed to explain the principles of digital color. This recently U.S.-patented device, aptly named the COLORCUBE, serves as a physical model of how color is stored, manipulated, and reproduced using digital processes.
Included with the COLORCUBE is a manual describing the 10 steps to understanding digital color. The following description is provided as an overview:
1) How the Human Eye sees Color
The eye contains two kinds of receptors: rods and cones. While the rods convey shades of gray, the cones allow the brain to perceive color hues. Of the three types of cones, the first is sensitive to red-orange light, the second to green light and the third to blue-violet light. When a single cone is stimulated, the brain perceives the corresponding color. That is, if our green cones are stimulated, we see "green". Or if our red-orange cones are stimulated, we see "red". If both our green and red-orange cones are simultaneously stimulated, our perception is yellow.
The eye cannot differentiate between spectral yellow, and some combination of red and green. The same effect accounts for our perception of cyan, magenta, and the other in-between spectral colors.
Because of this physiological response, the eye can be "fooled" into seeing the full range of visible colors through the proportionate adjustment of just three colors: red, green and blue.
Spectral Sensitivity Curve for each of the cones in the human eye.
2) Identifying Primary Colors
Any color can be spectrally analyzed using a prism to determine its red, green and blue primary values (additive color space), or its cyan, magenta and yellow primary values (subtractive color space). This simple yet powerfultechnique can be used to identify true primary colors. Choosing the correct three primaries maximizes the number of colors reproducible within a color space.
Viewing these circles through a prism reveals the primary colors. The circle on a white background breaks into Cyan/Magenta/Yellow primaries. The same circle on a black background breaks into Red/Green/Blue primaries.
3) Additive and Subtractive Color
Televisions, cameras, scanners and computer monitors are based on the additive system of color (RGB), where red, green and blue light projected together yield white. Offset printing, digital printing, paints, plastics, fabric and photographic prints are based on the subtractive system of color (CMY/CMYK) in which cyan, magenta and yellow mix to form black (K).
The unique feature of the COLORCUBE is that both systems are integrated within one model. Switching between RGB and CMY color systems is as simple as turning the model over.
RGB and CMY vertices, when placed in the same referential color space, form the outer dimensions of a cube.
4) Color Models
With each color theory advancement comes a new model by which to understand it. Unfortunately, users of older color technologies rarely, if ever, adopt these new models. For example, the color wheel is virtually identical in appearance and operation to how it was first conceived by Sir Isaac Newton. Painters continue to incorrectly define primary colors as red, yellow and blue according to the color wheel despite the fact that such technologies as offset printing and photography, each almost a century old, are based on a three-dimensional system of color using the true primaries cyan, magenta and yellow.
Other models of color used by specialists in their respective vocations include: Hue/ Saturation/Value (HSV), CMYK charts, RGB Color Space, Pantone color system, CIE Color Space, DIN color chips and spectral luminance graphs.
Computers and other digital devices define color based on a new model of color known as a COLORCUBE, which defines the digital representation of color.
5) Storing Images in Computers
All digital color devices that handle the storage, manipulation, and reproduction of color images do so by storing RGB values. Digitally storing an image requires that it first be broken down into a grid of tiny pixels (dots). Each pixel is sampled for the amount of red, green, and blue light present. The entire image is then stored one pixel at a time. To store a 3-inch square image at 150 dots per inch requires storing 202,500 pixels or 607,500 bytes.
The theoretical model describing how colors are stored in a computer is often displayed as a cube. This method of storing color has proven to be remarkably adaptive, allowing conversions to a wide variety of color models; including the color wheel, CIE color space, HSV color, Munsell Sphere, Pantone system, DIN chips, and spectral definitions of color.
The fundamental difference between the COLORCUBE and all other models of color is that it describes colors within a color space based on measured input quantities (what quantities of primary pigments are used to make the color). Other models of color are based on measured output values (what the color looks like). Basing a color system on input values considerably simplifies issues related to color naming, color reproduction, color visualization, color calibration, color manipulation, and color mapping between color spaces.
6) Visualizing a Color Space
The ability to visualize all the available colors within a three-dimensional color space and to see the inter-relationships between those colors is a huge advantage when working with color. Although there are a number of computer diagrams simulating a theoretical color space, the COLORCUBE model is the first of its kind to define a physical model with the interior colors visible.
As the eye can see over 16 million colors, the key to the COLORCUBE concept is that the external edge points of the cube are defined, and interior colors then approximate the range of colors between end-points. This then defines the outer dimensions of the visible color space, while allowing the viewer to see the internal elements. Color cubes of increasing density can then be generated based on a required "total" number of colors desired. A COLORCUBE that defines all colors reproducible within a color space would be 256 cubes on each side, for a total of 16,777,216 elements.
Planes of color in 3D color space.
7) Color Mixing
Each color element within a COLORCUBE has a unique numeric identifier indicating what proportionate input values were used to reproduce the color. Each element also has a unique position within the cube, thereby ensuring that one can easily map between positional information and mixing information. If the mixing information is given, then the positional information can be deduced. If the positional information is given, then the mixing information can be deduced. This feature of the COLORCUBE eliminates much of the guess work associated with naming, mixing, and describing a color, and ensures that within a defined color space, digital colors remain consistently reproducible.
8) Color Selection
The unique three-dimensional placement of colors within the COLORCUBE model works well as a color selection tool. Using the cube, it is easy to choose complementary colors, harmonious color runs, warm colors, cool colors, tints, shades, and colors of equal value. All color relationships can be shown to be mathematical in nature, and can be modeled using simple XYZ axis Cartesian math.
9) Color Manipulation
To manipulate colors within a color space one must first define a set of mathematical rules by which colors can be modified. Color Math, as it is referred to, relies upon first breaking colors into their constituent primary values, then doing the mathematical operation. The end result is mixing instructions for a new color that can be found in the COLORCUBE.
For example, to predict the result of adding two colors together, break down each color into its primary proportionate values. Then, add together the like primary values from both colors. The combined total for each primary yields the positional coordinates for the resulting color within the COLORCUBE. Similar logic can be applied to color subtraction (subtracting one color from the other), and to higher level operations such as adjusting contrast, brightness, and saturation.
Color math in subtractive color space:
10) Color Mapping and Calibration
The root of all color calibration and color mapping problems is that color spaces used by different color reproduction processes do not define the same visible area. Each color space is a subset of the true range of visible colors. To effectively map colors between different color spaces, a calculation must first be made to determine the color relationships between each of the color spaces. The objective when mapping colors is to find the best approximation so that the final image does not appear blatantly altered.
The current solution to correctly mapping colors between two color spaces requires spectrally analyzing the output characteristics of each device under controlled lighting conditions, and mapping the colors back to a CIE definition.
Color mapping models in such popular software programs as Corel Photo Paint, and Hewlett Packard Scanning software, provide two-dimensional color calibration interfaces which are difficult to use, are incomplete, and require sophisticated knowledge about color.
Software user interfaces that support color mapping could be vastly improved by recognizing the three-dimensional nature of color. Color space mappings could be visually represented in three-dimensional space relative to each other, and relative to the theoretical set of visible colors.
If a $50,000 digital color system does not perform to expectations, the end user is likely to conclude they need more training. However, if a $5,000 digital color system does not perform to expectations, the end user is likely to conclude the product is broken.
As digital color products become less expensive and sales volumes increase, the availability of economical product training will become an important issue. For users to understand how best to recognize and deal with complex color problems, they must become familiar with the fundamentals of digital color.
The COLORCUBE is an elegant model of digital color which can be used to teach simple color concepts. Users can learn and properly understand the basic physiology of color perception, the intricate relationship between additive and subtractive color systems, and the mathematics of color image manipulation.
At a time when art, science and other color-intensive industries are converging in the digital realm, a unified vision of color must emerge. The definitive model for that vision is the COLORCUBE.
|The COLORCUBE website is sponsored by ImageMAKER Development Inc.|
|http://www.ConversionServer.com||Convert Documents & PST files to TIFF, JPEG & PDF|
|http://www.Discovery-Assistant.com||Electronic Discovery & Litigation Support Software|
|http://one-way-lightspeed.com/||Measuring the One-Way Speed of Light|
|http://www.ImgMAKER.com||Windows Fax print driver|
|Home Site Map Order Feedback Copyright 2000-2018 © Spittin' Image Software, Inc. All rights reserved. Back to Top|