Appendix 4: An Introduction to the Applications of Color in Computer Graphics

[COLORREZ.RPT] [Began: 1989] [Printed: March 24, 2000]

"A long time ago in a galaxy far, far away . . . "[Lucas, 1977]. Thus begins the epic adventure that launched computer graphics into international acclaim. George Lucas captured the thrill of adventure using special effects and computer graphics to dazzle enchanted viewers. The company he formed to produce Star Wars went on to become the leading producer of films that use computer graphics [Industrial Light and Majic (ILM)].

Graphics were needed to make the data computers manipulate (produced) more accessible to the human mind. Computers use (create) immense amounts of data. Computer graphics help humans understand the data as information. Color was needed to enhance this process. Color makes the graphics more appealing, both because the data is displayed as information (more legible), and for psychological reasons that are not completely understood to date. Today, computers can produce graphics as immense and complicated as the data. Current research is in writing programs to limit the necessary data used to model the data as graphics (composing the graphics).

Computer graphics were developed in the mid 60's, "a long time ago" to a computer, since these machines split the second into billionths to count the passage of time (nanoseconds). Computer graphics began with Ivan Sutherland, who developed the first program to allow a user to interactively draw graphics on the computer screen [paintbox?]. Marvelous as this process was, within a decade, Jim Blinn analyzed such things as the colors of the spectrum and how light is reflected and refracted on the surfaces of various shapes of polygons and coded it into computer programs. The value of Jim Blinn's contributions was epitomized when, Ivan Sutherland said, "There are only a dozen good people in computer graphics, and Jim Blinn is six of them"[Rivlin, 1986]. The pioneers wrote the programs that enable mere mortals to apply the rules of physics they encoded and produce computer graphics with relative ease.

With access to the computer code, programmers have written many types of graphics programs that have spread the use of computer graphics to all walks of life. From forecasts presented in the daily paper to forecasts and network logo promotions on television to entertaining films seen at movie theaters, we are inundated with computer graphics [Weather Data Inc., 1989][Fuller, 1989][Lisberger, 1982].

The graphics are a result of the computer's ability to manipulate large quantities of numbers. They resulted from the user's need to view these numbers in a form that was easier to understand. "Computers greatly expanded the volume of data that could be handled, and graphical displays eased the chore of understanding data by presenting relationships visually." [Mundie, 1989] As the computer evolved from a fast calculating machine to an omnipotent "workstation," the basic premise of its existence has not changed: to manipulate massive amounts of numbers at a speed and accuracy beyond human capabilities. What has changed is the output device (interface) that the computer operator uses to check the results of commands given the machine. From a row of eight lights that blinked the input and output in binary code, necessity created the hi-resolution color cathode ray tube used to monitor the activity and results of modern computers.

It was discovered that color coding of the information was a benefit to finding the correct piece of information [Rogers & Earnshaw, 1987]. As tremendous amounts of data are produced daily, color graphics will be a tool to help find, understand, and use the data as information. Perhaps the future will be as Gibson predicts: "People jacked in so they could hustle. Put the trodes (electrodes) on and they were out there, all the data in the world stacked up like one big neon city, so you could cruise around and have a kind of grip on it, visually anyway, because if you didn't, it was too complicated, trying to find your way to a particular piece of data you needed. Iconics, Gentry called that."[1988]. An "icon" is a simple type of computer graphics, an object on the screen representing a function that can be performed. It is becoming commonplace to use icons to make it easier for the user to operate the computer without remembering (knowing) precise commands.

Not only are the commands given the computer being represented with these graphical icons, but the results of the computation itself are transformed into graphs or pictures. As these graphics have become more complicated, monochrome graphics have no longer been sufficient to represent the level of detail needed. The spectrum of colors is now required to differentiate separate entities and enhance the relationships of these entities one with another. Monochrome graphics were limited to on/off pixels (variations of brightness) to differentiate between objects. Color, by definition, adds hue, saturation, and brightness as graphic variables. "Color is the attribute of visual experience that can be described as having quantitively specifiable dimensions of hues, saturation, and brightness"[Burnham, Hanes, & Bartleson, 1963]. Introducing color to computer graphics solved some problems, but created many new ones. Color gave computer graphics many new choices, but the effect of each color choice varied widely and required study.

The color used to depict any object takes on one or more of three roles each time a computer programmer chooses a color. Objects that have been modeled previously require "Established" usage of color. "Realistic" colors are chosen to represent an object as we are accustomed to seeing it, and to represent objects not visible in expected "realistic" color. "Psychological" implications affect color choice when selling a product, setting up a mood, or considering esthetics. Two of these roles could be used in conjunction to "try out" new ideas or theories.

The color choices of realistic and established usage were set. It was the psychological aspect that prompted research into how humans react to the colors they view [Cornsweet, 1970]. Individuals prefer colors because of the type of experience they link to a color. There have been numerous tests to determine the preferences of single colors, which are, in order of preference: blue, red, green, violet, orange, and yellow [Burnham, Hanes, & Bartleson, 1963]. Color preference is one of the many factors that affect the use of color in computer graphics. Popular software such as WordPerfect, Wordstar, and Turbo Pascal use blue as the default color, but allow the individual user to change the color. Even such a simple decision as what color to make the background (backdrop) of a graph can be affected by color preference.

Solid displays of color used as backgrounds (backdrops) to a graph will influence the viewer, but there is more to color selection than just the background color. A complete list of the effects of color is at first overwhelming, but, taken separately, the visual impact of color can be determined. Michael Wilcox determined the interactions of color to be separable into these groups: Complementaries, simultaneous contrast, border contrast, vibrating color, two color harmony, optical mixing, warm and cool colors, advancing and receding colors, colors from black and white, color memory, and the psychological effects of color [1983]. When there are this many choices to make, each affecting the others, the mathematical complexity is obvious (considering the implications of color use and the availability of 16 million colors). Most of these interactions are impacted in the psychological effects, so these colors must be viewed in light of their psychological shading.

The mood a viewer receives from a particular color seems to contradict some of the order of preference of solid colors. To the viewer, blue is reliable, calming, cool, and creates a feeling of security and tranquility. Red is argumentative. Green is soothing, although certain brighter shades can bring on nauseous feeling. Violet is unsettling, because the viewer cannot decide if this is a red or a blue color. Orange is acceptable, expansive, outgoing and warm. Yellow is bright and cheerful. Brown is easy to live with and practical [Wilcox, 1983].

Besides mood, Color has weight, size, space, and "memory". The brain's reception of color relates not just to how it thinks of the color received but, given an object, the brain "remembers" the color of the object in natural light and will change the eye's perception of the color of these objects in dark or unnatural lighting. The term for this phenonomen is the "memory color" of the object [Katz, 1935]. Another psychological factor of color perception is brightness, since brighter colors seem closer than dull or dark colors. Darker colors seem heavier; therefore, they are placed at the bottom of the screen where they seem to belong. The only specific color which has attributes of size and space is red. Red objects appear closer and red spaces appear smaller [Wilcox, 1983]. Whether by accident or design, all these psychological effects of color have an effect on the viewer.

The audience certainly enjoys the colors and objects displayed on the screen, but what is actually represented is important. The graphic representation of objects on a computer screen is formed by these five basic graphic elements: (point, line, polygon)(solids) polyline, polymarker, text, fill area, and cell array [Hopgood, Duce, Gallop, & Sutcliffe, 1986]. A polyline is a line, a polymarker is a dot, and text is simply that. Fill areas are polylines connected and filled in to represent areas of an object's surface. Cell arrays are used when the image is actually a matrix of dots that represent differences in the perception of an object as viewed through a camera or other imaging device. Cell arrays are useful in satellite images like Landsat, where each dot represents 30 meters of the Earths' surface in values of 1 to 256 broken into different portions of the spectrum. Using these basic graphic elements, all objects can be represented. One of the amazing abilities of computer graphics is to create an image of invisible objects, whether they are too small to be seen by the human eye, too large to see all at once, or don't exist.

This ability to depict "invisible" objects is used in computer-aided design (CAD), computer-aided manufacturing (CAM), computer graphics, computer art, animation, simulation, computer vision, robotics and artificial intelligence [Mortenson, 1985]. In the modern world computers produce images that will let our minds see what was not comprehensible in scientific visualizations, computer graphics in medicine, flight simulators, art and entertainment [Ward, 1989]. Another application of computer graphics is Geographic Information Systems (GIS) which portrays maps (polylines) on computers with overlays of satellite imagery (cell arrays) and links to a database of information (text) about the individual objects (polypoints) of the map that the screen displays.

Keeping track of all the different color parameters, billions of bits of data, while trying to define a graphic representation could be a formidable task. Though best left to a computer, it can still take days for the fastest computer to formulate a complete graphic image. To reduce the time required to formulate a computer image from data, the first step is to determine what it is in the problem that actually affects the solution. That is, what data is needed? Then, the computer must be told, via the program, to create a graphic which only depicts those parameters. "Research now centers on deciding the requirements to model an object correctly, not completely but concerning the aspects of the problem"[Marvin, Labovitz, & Wolfe, 1987]. Reducing the parameters also reduces the number of colors required to portray the image. The number of colors available can limit the capability to portray shading on objects. Computers are capable of producing graphics of great detail, but the computing time and colors needed can be prohibitively expensive.

For some applications, computer graphics need to simplify the image to wire frame models, molecular models, or components of a bridge stress test. For other applications, such as computer movies, architectural rendering, and scientific simulations, the image requires a photorealistic quality. This requirement led to further specifications to define the image. The lens of the camera and the eye record blurred or out of focus images and photorealistic quality in computer graphics would duplicate this, as well as "temporal filtering so that the opening and closing of the shutter causes moving objects to be blurred, and depth of field so that only objects at the current focal distance are sharply in focus" [Pixar, 1988].

These photorealistic qualities do not require any additional graphic elements. The effect of photorealism is brought about through intensive modeling procedures written into the program. "In the future, computer graphics may replace the photographic medium as a means of communication with images" [Rivlin, 1986]. The quest for photorealism has spawned new products. Kodaks' new Digital camera can input images into the computer. Black and white images (movies) can be "colorized" image by image. In the future, the same process will allow the image of another actor or the home viewer to replace the original actor, or historical figure. The home viewer will have a choice of colors that would be colorized into the decorating scheme of any image viewed on the television.

Color computer graphics can be as simple as a businessman's bar graph or as complicated as a visualization showing the stress of a bending rod in a spectrum of colors relating to the amount of stress. Just as Jim Blinn coded the physics of color, in the future, another computer programmer might encode the psychological effects of color, so that the color of computer graphics could have the desired psychological impact. Today, many people still fear computers as a monster to be tamed. Perhaps color computer graphics, by making it easier to find, understand, and use data, can make this monster not only friendly, but a willing mediator with the machine world.

Artwick, B.A. (1984). Applied concepts in microcomputer graphics. Englewood Cliffs, NJ.: Prentice-Hall, Inc.

Burnham,R.W., Hanes, R.M., & Bartleson, C.J. (1963). Color: a guide to basic facts and concepts Vol. 20 Basic elements of color education. New York: Wiley.

Cavanagh, P. (1987). Reconstructing the third dimension: Interactions between color, texture, motion, binocular disparity, and shape. Computer Vision, Graphics, and Image Processing, 37(1), 171-195.

Cornsweet, T.N. (1970) Visual perception. New York: Academic Press.

Fuller, J. (Meterologist). (1989, May 21). Eyewitness news. St. Louis: CBS.

Gibson, W. (1988). Mona lisa overdrive. New York: Bantam Books.

Hopgood, F.R.A., Duce, D.A., Gallop, J.R., & Sutcliffe, D.C. (1986). Introduction to the Graphical Kernel System (GKS) Vol. 28 A.P.I.C. studies in data processing. Orlando: Academic Press.

Katz, D. (1935). The world of colour. London: Kegan Paul, Trench, Trubner & CO. LTD.

Lisberger, S. Director. (1982). Tron [film]. Los Angeles: Walt Disney.

Marvin, J.W., Labovitz, M.L., & Wolfe, R.E. (1987) Derivation of a fast algorithm to account for distortions due to terrain in Earth-viewing satellite sensor images. IEEE Transactions on Geoscience and Remote Sensing, GE-25, 244-251.

Meyer, G.W. (1988). Wavelength selection for synthetic image generation. Computer Vision, Graphics, and Image Processing, 41(1), 57-79.

Mortenson, M.E. (1985). Geometric modeling. New York: John Wiley & Sons, Inc.

Mundie, C. (1989, April). Interacting with the tiny and the immense. Byte, pp. 279, 280, 282-284, 286, 288.

Pixar. (1988). The Renderman interface (Version 3.0). San Rafael, Ca.: Author.

Ranson, K.J., and Daughtry, C.S.T. (1987) Scene shadow effects on multispectral response. IEEE Transactions on Geoscience and Remote Sensing, GE-25, 502-509.

Rivlin, R. (1986). The algorithmic image. Redmond, WA.: Microsoft Press.

Rogers, D.F., Earnshaw, R.A. (Ed.). (1987). Techniques for computer graphics. New York: Springer-Verlag.

Savoy, R.L. (1987). Contingent after effects and isoluminance: Psychophysical evidence for separation of color, orientation, and motion. Computer Vision, Graphics, and Image Processing, 37(1), 3-19.

Ward, F. (1989). Images for the computer age. National Geographic, 175, 718-751.

Weather Data Inc. (1989, May 21). Weather. St. Louis Post-Dispatch, sec. A, p. 2.

Wilcox, M. (1983). Color theory for oil colors or acrylics. New York: Watson-Guptill Publications.