Color as a Determined Communication

7/30/2019 Color as a Determined Communication

1/13

Copyright 1996 by International Business Machines Corpora-tion. Copying in printed form for private use is permitted withoutpayment of royalty provided that (1) each reproduction is donewithout alteration and (2) theJournal reference and IBM copyrightnotice are included on the first page. The title and abstract, but noother portions, of this paper may be copied or distributed royaltyfree without further permission by computer-based and other infor-mation-service systems. Permission to republish any other portionof this paper must be obtained from the Editor.

JACOBSON AND BENDER 0018-8670/96/$5.00 1996 IBM IBM SYSTEMS JOURNAL, VOL 35, NOS 3&4, 1996526

Although it is possible that one viewersperception of color may be very different fromanothers, experimental evidence suggests thatthe relationships between colors are, in manyrespects, universal, and thus relatively freefrom individual and cultural influences. Theexperience of color can be describedobjectively, so that predictable visual sensationscan be elicited by adjusting the relationshipsamong colors. A model of color experience isdescribed that is based on the types ofinteractions among colors. The model adjustsformal compositional attributes such as hue,value, chroma, and their contrasts, as well as sizeand proportion. Components such as these can beutilized to build a general architecture for addingguidance to interactive systems.

There was gold paint, but Rembrandt didnt use it to painta golden helmet.

Wittgenstein

n engineer or scientist would use gold paint topaint a gold helmet. Why did Rembrandt choose

not to use gold paint? The engineers or scientistsquantitative understanding of color is far removedfrom an artists qualitative understanding of color.

The former considers color as something that is speci-fied or measured in terms of metrics such as nanome-ters or just-noticeable-differences and speaks of colorin terms of detection, discrimination, legibility, andcontrast. However, for the artist or the poet, color issomething to experience, not measure and quantify.Our current understanding of human color visionemanates from the traditions of both artists and scien-tists. Newton characterized light; Goethe contem-plated its appearance.

In human visual experience, colors appear as interre-lated sensations that cannot be predicted from theresponse generated from viewing colors in isolation.People can make consistentevaluations of the magni-tude of any given experience of colors based on thetype of interaction among colors. People respond tothe relationships among colors.

Color experience is governed by well-defined objec-tive principles that can be quantified. These principlesare applicable to a wide variety of disciplines. Forinstance, in interface design, color can reinforce infor-mation by providing a visual counterpoint. In imagereproduction, color matching becomes a matter ofpreserving the experience of color. In graphicdesign, a wide variety of visual experiences can beestablished and transposed. In multimedia applica-tions, sensations produced by different modalities canbe transcoded or integrated.

The common ground between the measurement ofcolor and the application of color to design, thehuman experience of color, is found in the interactionbetween analytical and empirical data. In this paper,the search for a common ground begins with a quali-tative description of color, followed by a review ofquantitative descriptions of color. An experiment that

models color expressiveness is described. The paperconcludes with rules for creating a language ofcolor

A

Color as a determined

communication

by N. JacobsonW. Bender


2/13

IBM SYSTEMS JOURNAL, VOL 35, NOS 3&4, 1996 JACOBSON AND BENDER 527

Figure 1 From black and white to color (photograph of moon courtesy of NASA): (A) a lunar landscape composed fromgray, (B) the elements used to create the landscape, (C) the addition of yellow, (D) the addition of more colors,

and (E) the cacophony of too much color

Figure 2 A sense of order: a succession of masks is applied to the cacophony of colors from Figure 1E in order toextract (A) monochrome, (B) analogous, (C) beyond analogous, and (D) complementary relationships

(A)

(B) (C)

(D) (E)

(A)

(B)

(C)

(D)


3/13

JACOBSON AND BENDER IBM SYSTEMS JOURNAL, VOL 35, NOS 3&4, 1996528

communication and how these rules may be applied todesign systems.

The quality of color

How does color contribute to communication? Inorder to answer this question, first imagine a worldwithout color, an achromatic world that consists ofblack, white, and shades of gray but no other colors.People lacking photopic vision live in such a world.Others, such as the man who is the subject of Dr.Oliver Sackss essay entitled The Case of the Color-Blind Painter,1 see the world without color becauseof a visual cortex injury (cerebral dyschromtopsia).Sackss patient lost the ability to distinguish betweenhues but otherwise retained the ability to see. Thepatient saw objects, facial features, silhouettes,dynamics, and depth and focus, but could not seecolor. How was his visual message-processing systemimpaired by his inability to see color? How was hisability to communicate visually affected both quanti-tatively and qualitatively?

An examination of a monochrome image providesinsight into the world of the color-blind artist. In Fig-ure 1A, the moon is reproduced as a gray-scale image.Surface texture provides a sense of the lunar terrain.Shape and volume are evident from the shading orchiaroscuro. The relativeposition of the craters can be

determined. Changes in the craters size and orienta-tion across the surface impart a sense that they arereceding toward the horizon. And, attention is drawnto the crater Copernicus on the horizon, in part,because of the relatively large contrast of valuebetween the sunlit sides and the shadow across itsbasin.

Some of the color elements used to create this repro-duction of the moon have been extracted and areshown in Figure 1B. They are shown as patches ofgray. Figure 1A is composed from these achromaticelements. Together these patches create an orderedprogression from dark to light.

This black and white view of the world is similar tothat experienced by the color-blind painter. Althoughrich in detail, it is void of color, and thus void of theexpressive qualities of color. Consider what happenswith the addition of color: Yellow has been added toFigure 1C; it changes the character of the message byadding another dimension to the composition. Theaddition of color stimulates and excites. As morecolor is added, attention is drawn away from the gray

patches (Figure 1D), and a new sense of orderemerges: a progression of hues.

However, with the addition of still more color, a pointis reached where a consistent message can no longerbe viewed in the image (Figure 1E). What began as anordered progression has turned into a cacophony ofcolors. Whatever color relationships existed aremasked by the complexity of the image. Attention isdrawn to too many places at once because of incom-patibilities in the ordering or segmentation of theimage elements.

A sense of order. A middle ground can be foundbetween the two extremes of no color and too muchcolor. By application of masks designed to exploit

selected dimensions of color appearance, order can beextracted from the chaos of Figure 1E. Figure 2Aillustrates one such mask. All of the colors revealedshare a common hue (in this case, yellow). Althoughyellow ranges in value from dark to light and inchroma from gray to highly saturated, the colorsbelong together, as part of a stable, ordered family.This organization of color is commonly referred to asmonochrome.

Figure 2B illustrates another selection of colorsbrought out by the use of a different mask. A progres-

sion ofanalogous (or neighboring) hues following theorder of the spectrum is extracted. The order found inthis progression is familiar to all with normal colorvision. Although more lively than the monochromepalette, it, too, is a stable arrangement.

A departure from the analogous ordering of hues isillustrated in Figure 2C. The mask reveals a progres-sion from yellows to greens to blues. This ordering ofcolors is more stimulating but also more stressful and

Expression is used todescribe the human

responseto appearance.


4/13


ambiguous. This arrangement, stretching beyondanalogous, is less balanced than the monochrome or

analogous selections.

Figure 2D illustrates yet another ordering. The maskexposes complementary colors (colors that intermixto gray). Although the complementary relationship iseven further beyond analogous than the progressionshown in Figure 2C, there is no visual tension orambiguity generated from the colors. There is a senseof a dynamic energy associated with complementarypairings. Whereas the analogous progression of huesand the monochrome field are ordered and stable, thecomplementary progression invokes images of adance.

For those of us with normal color vision, the vocabu-lary of color communication is something both innateand intuitive. We do not rely on prescriptions of colorsemantics or rules of color harmony to determine aresponse to color. Our response to gravity and oursense of balance are based on an internal plumbline,not Newtons equations. Similarly, our response tovisual sensations and our sense of dynamics and stressare inherent. It is this internal color plumbline thatSackss patient lost.

A hierarchy of color communication

Historically, color has been characterized by consider-ation of its application. Considerations have includedthe physics of color, the mechanisms of the humanvisual system (HVS), application to coding and repro-duction, and application to design and interactivity.Color has been modeled as a physical, psychophysi-cal, psychological, and metaphysical phenomenon.Evanss book, An Introduction to Color,2 providesgood background on the subject. Over the past fortyyears, through both clinical and neurophysiologicalstudies, models of the HVS response to color havebeen developed, beginning with the detection of radi-ation in the retina and continuing as far as the visualcortex. Zeki3 and Hubel4 have written detailed

descriptions of their hypotheses of the cortical pro-cessing of color.

The discussion of color in this paper is organized as ahierarchy (Figure 3). At the bottom level is the detec-tion of points of electromagnetic radiation. At the nextlevel is the appearance of radiation as an ensemble ofpoints. At the top level are the inferences made fromappearance. These inferences are the building blocksof color communication.

The HVS responds to electromagnetic energy betweenapproximately 370 and 730 nanometers. The conceptof visible radiation is used principally to describeenergy external to our bodies and is defined, dis-cussed, and applied using the language of physics.Radiation can be modeled as a collection of stimuli,each independent of its neighbor. A description ofradiation is most often a description of points. Use ofthe word radiation is restricted to describing phenom-ena by wavelength, phase, and amplitude at singularpoints or across uniform fields. Examples of radiation

models include that of Newton, who, by using prisms,established the principles of color mixing. Pantone**,a popular color atlas used by the printing industry, is acollection of radiation samples organized by the tintsand shades of pigment.

A goal of color science is to model the transformationof radiation by the HVS into the sensation of color.Many models are designed to predict color appear-ance by scaling radiation based upon parameters

Figure 3 A hierarchy of color communication is aprogression from the detection of radiation,

through the perception of appearance, to theinferences of communication.

COMMUNICATION

APPEARANCE

RADIATION

CO

NTEXT

POINT

E ( )R ( )( )d

370

730

MUNSELL

LETTVIN

JACOBSONGOETHE

NEWTONDETECTION

INFERENCE

MAXWELL

KOBAYASHIMOON & SPENCERSIVIK & HRD

HUNT

CIEHERING

PANTONE

LAND

NCS


5/13


found in human psychophysics. Examples includethat of the Commission International de lEclairage

(CIE), Munsell**5,6 (Figure 4), and the Natural ColorSystem (NCS)**7 (Figure 5). These models vary interms of their support for the independence of particu-lar primary colors, mathematical complexity, andvisual ordering. The latter is a particular strength ofboth the Munsell and NCS systems. Both of these sys-tems are perceptually uniform and orthogonal (Figure6). These models are useful for tasks such as compar-ing different color devices, predicting the results ofcolor mixtures, and finding an approximate comple-ment.

A limitation of these models is that they largelyignore visual context, i.e., they do not predict how the

appearance of a color changes depending upon itsproximally surrounding colors. This visual contextcan produce large shifts in the perception of a colorthat cannot be accounted for by colorimetric specifi-cations of isolated points of color, since the appear-ance of color is the result of an interaction of colors(Figure 7).

The concept of color appearance is used to describethe retinal and cortical responses of the visual systemto radiation. Color appearance is not an attribute ofobjects external to an individuals body. However, it isan attribute of representations within an individualsmind; hence, appearance is defined, discussed, andapplied using the language of perception. Humanresponse to radiation is derived from internal process-ing of the relationships between points of radiation.Examples of appearance models include those devel-oped by Land,9 Lettvin,10 and Hunt.11

Color appearance models, just like colorimetric mod-els, ultimately describe the phenomenology of a sin-gle color. That is, they describe the processes bywhich the appearance of a single color is affected byits surrounding colors, whether it is due to contrasteffects, induction, assimilation, adaptation, or spatialor temporal configuration.

Wandell,12 in Foundations of Vision, defines seeingas the process of deriving meaning from a collec-tion of visual inferences about the world. Multipleinferences are reconciled to create a unified explana-tion of the stimuli. In keeping with Wandells defini-tion, seeing color involves more than just thesensation of isolated color appearances. It is the resultof a plurality of colors taken as an ensemble, indepen-dent of how the appearance of the individual colors is

established. It is only within the context of an ensem-ble of colors that color conveys semantic and sym-bolic information. Communication tasks such asnaming, classifying, searching, quantifying, ordering,highlighting, and conveying expression or emotionare consistently determined by the use of a plurality ofcolor appearances.

The word expression is used to describe the humanresponse to an ensemble of color appearances. Wehave an internal psychophysical response that is initi-ated when viewing multiple external radiations.Although it is possible for one viewers perception ofcolor to be very different from anothers, experimentalevidence suggests that the relationships among inter-nally generated colors are, in many respects, univer-sal, and thus relatively free from individual and

Figure 4 Munsells Book of Colorsis organized in termsof three perceptual attributes of color: hue,

chroma (saturation), and value (lightness). (Atable converting colors from Munsell to CIE

can be found in Wyszecki and Stiles.6)

Figure 5 The Natural Color System is based uponHerings Opponent Color Theory. (A) Thesystem takes as its basic elements: black,white, red, green, blue, and yellow. Thesecolors are taken as absolute mentalreferences, from which all other colors aredetermined by their degree of resemblance.(B) Grays are described by the percentage ofblack and the percentage of white theycontain. (C) Chromatic colors are described interms of all six components.

HUE

VALUE

CHROMA

W

S

C

Y

G

B

R

(A) (B) (C)

C


6/13


cultural influences. Examples of qualitative models ofexpression are Jacobsons Relative Harmony Quot-ient13 and Kobayashis Scale of Colors.14 Examples ofquantitative models are Moon and Spencers Geomet-ric Formulation of Harmony15 and Sivik and HrdsSemantic Dimensions.16

Robust communication. In order to inform, per-suade, and stimulate (or fatigue), messages to thevisual system must have integrity, credibility, andexcitement (or lack of excitement). Making the mes-sage available is sufficient for some applications, butfor most, the message must be both available and easyto understand.

To design a message that is distinct, reliable, anddraws the attention of the observer, one must evaluatethe degree of visual prominence of the message. Thismeans determining whether the message is able to beperceived and decoded. It also needs to have its properplace in the hierarchy of information so that it attractsthe attention it deserves. Congruence among mes-sages can make them easier to receive; incongruencecan lead to the inhibition of communication, or mis-communication (Figure 8).

Figure 7 Colorimetric models can predict whendifferent radiations, such as (A) and (B), will

match, but not how they will appear. The samegreen appears different in (C), (D), (E), and (F)due to differing contexts. Appearance modelscan predict what colors in context will looklike, but not what they communicate. (G) Aninteresting exception to appearance modeling

is Adelsons illusion:8 The diamond shapes inthe light column in the center of the figureare the same as the diamond shapes in thedark columns on either side.

(A)

(B)

(C)

(D)

(E)

(F)

(G)

370 730

370 730

Figure 6 A perceptually uniform and orthogonalordering facilitates visual navigation: (A) when

saturation extends to 100 percent at all valuelevels, circuitous paths must be used toindependently change value and saturation,but (B) when value and saturation increase inperceptually uniform increments, the paths arestraight.

(A)

(B)


7/13


We need to grade the efficacy of various means ofgenerating messages for the tasks we want to perform.

Grading can lead to design methodologies for creatingmessages that are appropriate for our applications.

Experience of color. In visual communication, a widevariety of color experiences can be established. Pat-terns of color combinations constitute experiences ofcolor, each with its own unique character. However,even among varied experiences, it is possible to iden-tify features that are common among color experi-ences: some experiences are energetic and others areweak.

Humans make consistent evaluations as to where any

given color experience, produced by interactionsbetween two colors in patterns encompassing a widefield of view, lies in a range of magnitude. We have anability to make judgments about the magnitude of theinteraction of colors based on how colors relate to oneanother. Features of the colors, such as chromaticcomposition and spatial configuration, determine themagnitude of the interactions. These features consti-tute the dimensions of color experience.17

A study by Feldman, Jacobson, and Bender18 foundthat when evaluating the experience of color, agree-ment among people is strong; that is, judgments of themagnitude of experience constitute an invariant aspect

of human response to color. Beyond such invariantresponses lies a more personal and subjective issue of

judgment, such as whether an experience is pleas-ing. The problem with assessing personal prefer-ences is that there is no consensus among people,even less among different cultures, as to which colorsare beautiful or ugly together. Therefore, subjec-tive evaluation of color is deliberately avoided. Acomplementary study by Green-Armytage19 evaluatesthe role of personal preferences on the color experi-ence.

Architectures for interaction. Todays application

and development environments, although increasinglycapable of providing the user with the ability to inter-actively select colors (as well as fonts and othergraphical elements), do not offer the user guidancewhen making these selections. Consequently, thoughthe capacity for creating powerful visual communica-tions with color exists, it is often not realized since theuser is free to produce illegible and jarring results.Any architecture for interaction with color must con-sider radiation, appearance, and expression.

The remainder of this paper discusses how color expe-riences are established, described, quantified, andfinally, may be utilized in an interactive system.

Dimensions of color experience

Color is colors, plural.J. Albers

The experience of color is described with the directlyobservable features of the colors such as the chro-

matic dimensions of hue, value, chroma, and theircontrasts, as well as the spatial dimensions of size andproportion. Other features not considered here includefacial features, silhouettes, dynamics, and depth andfocus. The interactions between the dimensions wereuncovered through a series of experiments, which arereviewed briefly in this paper and detailed in Feld-man, Jacobson, and Bender.18 Semantic differentialscales, as defined by Osgood et al.,20 were used toevaluate systematic variations in the relationships

Figure 8 Color can hinder or help search: (A) colorinterferes with the task of finding the circle,

and (B) color, as well as shape, size, andposition, aids in the task of finding nominalchanges.

(A)

(B)


8/13


between the colors. Two of the scales of Osgood et al.were used: activity (tame-wild, still-vibrant) andpotency (quiet-loud). Their evaluation scale (pleasant-unpleasant) was not used since a nonjudgmental eval-uation was sought. It was found that distinct colorexperiences can be described with a single scale ofmagnitude so that seemingly disparate visual sensa-tions can be made commensurate with each other. Itwas also found in Jacobson, Bender, and Burling thatwithin each isobar of magnitudelow, medium, andhighdistinct qualities of expression are evoked,which vary from stressful to harmonic.21 Experienceof color is determined by several factors: the numberof colors, the brightness, the intensity, the size andshape of colored regions, etc. Thus, when evaluating a

color experience, it is not any one factor that is takeninto consideration but rather the experience of theinteraction between factors that is recorded. There-fore, to describe the experience of color, several fac-tors need to be considered, as well as their inter-actions.

Chang and Carroll22 used multidimensional scaling(MDS) to analyze data obtained from a color similar-ity experiment. Their analysis suggested that in addi-tion to Herings three color dimensions, lightness,red-green, and yellow-blue, there are folded dimen-sions. These folded dimensions approximate valuecontrast and hue contrast (as described below). Chang

and Carroll also found an additional dimension, splityellow, which they attribute to anomalies in some oftheir subjects. A more likely explanation of this arti-fact is the choice of sample points used in their exper-iment.

The dimensions of color experience are grouped intotwo categories: chromatic and spatial. The dimensionsused in the experiment designed by Feldman, Jacob-son, and Bender were:

Table 1 Seven dimensions of color experience and theireffect on response along semantic differential

scales

Dimension Effect onResponse

To IncreaseResponse

Reference hueHue contrastReference valueValue contrastChromaBlock sizeArea ratio

NoYesYesYesYesYesNo

Increase hue contrastRaise valueIncrease value contrastIncrease chromaDecrease block size

Figure 9 Seven dimensions of color experience

(A) REFERENCE HUE

(B) HUE CONTRAST

(C) CHROMA

(D) REFERENCE VALUE

(E) VALUE CONTRAST

(F) BLOCK SIZE

(G) AREA RATIO


9/13


Figure 12 Interaction between hue contrast and valuecontrast: value contrast increases from

bottom to top and hue contrast increasesfrom left to right. While the interactionincreases moving up and to the right, theregion of maximum interaction is indistinct.

Figure 13 The measured response of the interactionbetween hue contrast and value contrast

7

6

5

4

3

2

1

0 10 20 30 40 50 40 30 20 10 0

HUE CONTRAST

EXPERIMENTAL

2.00.5

2.5

0.0

MUNSELL VALUE

RESPONSE

CONTRAST

(IN MUNSELL HUE STEPS)

Figure 10 Interaction between hue contrast andchroma: chroma increases from bottom to

top and hue contrast increases from left toright. The interaction increases moving upand to the right. The region of maximuminteraction is in the upper right.

Figure 11 The measured response of the interactionbetween hue contrast and chroma

7

6

5

4

3

2

1

0 10 20 30 40 50 40 30 20 10 0

HUE CONTRAST

EXPERIMENTAL

maximum6.04.02.0

MUNSELL CHROMA

RESPONSE

(IN MUNSELL HUE STEPS)


10/13


Reference hue: When comparing two hues, one ofthe hues is designated as the reference hue, the other

as the related hue. In this investigation, the ten basicMunsell hues were selected as the reference hues(Figure 9A).

Hue contrast: Hue contrast corresponds to the align-ment between hues. It is defined as the distance inhue angle between any two colors, i.e., the amountof change in Munsell hue units between two hues.Hue contrast is a measure that indicates relativeangular distance from a reference hue, independentof the reference hue (Figure 9B).

Chroma: Chroma indicates the degree of departureof a color from a neutral of the same value. Chromacorresponds to the distance from the achromaticcore. In Munsell notation, neutrals are defined as

having chroma equal to zero. The chroma of a colorincreases up to a maximum number that varies,depending on the hue and value of the color (Figure9C).

Reference value: The Munsell color system can bevisualized as consisting of planes of samples ofequal value that are stacked perpendicular to a cen-tral achromatic core. When comparing two colors,the value of one color is designated as the referencevalue. With respect to the reference value, therelated color may be of either equal, higher, orlower value (Figure 9D).

Value contrast: Value contrast indicates the differ-ence in lightness between colors. It is computed by

measuring the distance in value between colors. InMunsell notation, value contrast is a number rang-ing between zero and ten units (Figure 9E).

Block size: Colors may be assembled in any sizeand configuration. For simplicity, in this investiga-tion colors were confined to two-dimensionalsquare waves whose characteristic frequencies var-ied in octave increments. These waves werearranged in a matrix configuration. The characteris-tic wave-length is measured in degrees of visualangle. The size of the blocks was large enough to bewell beyond the range of optical mixing that occursfor sizes less than 0.25 degrees. The blocks spannedangles of greater than 0.5 degrees. The block sizewas inversely proportional to the border (or shore-line) shared by the colors in the stimulus (Figure9F).

Area ratio: The number of blocks assigned to eachcolor within the stimulus varied. The ratio of thearea of the colors varied between 1:1 and 5:1 (Fig-ure 9G).

The experiment measured the interaction of theseseven dimensions of color experience. The magnitude

of response along the semantic differential scales to

hue contrast is used as a reference in the analysis ofthe data. The measured responses are bilaterally sym-metric around the reference hue.

The lowest values existed at the endpoints that corre-sponded to monochrome, and the peak values corre-sponded to complementary colors. Figures 10 and 11illustrate the interaction between hue contrast andchroma. Both the mean response and the spreadbetween minimum and maximum response are

Figure 14 Plot of Equation 1 showing (A) holding h1=Y,and (B) holding h1=BP

Figure 15 Rules of relative contrast are used toprescribe (A) quiet and (B) loud colors.

(A) (B)

(A) (B)QUIET LOUD

HUE - LOWVALUE - LOWCHROMA - LOW

HUE - HIGHVALUE - LOWCHROMA - HIGH


11/13


directly proportional to chroma. Figures 12 and 13

illustrate the interaction between hue contrast andvalue contrast. The mean response is directly propor-tional to the value contrast. The spread is inverselyproportional to value contrast. As value contrast var-ies, mean and spread vary in opposite directions.Graphically, as value contrast increases, the ends ofthe curves move up, resulting in a flatter response.

As value contrast increases, the flat response indicatesthat the effect of hue alignment is weakened because

of diminished spread. This suggests a strong interac-tion between value contrast and hue alignment.

The seven dimensions listed earlier are sufficient todescribe the phenomenology of color experience forsimple color configurations. However, additionaldimensions, such as separation between colors, enclo-sure of colors, and temporal effects, can be added.The results of numerous repetitions of the experi-ments indicated that five of the seven dimensions hada statistically significant effect on experience. Theyare hue contrast, value contrast, reference value,chroma, and block size. The remaining two dimen-sions, area ratio and reference hue, had no significanteffect. The overall results are summarized in Table 1.

The experimental data can be modeled by Equation 1:

f(v, w,x,y,z) = ac(0.5 x)2 + bc (1)

where v is the average value ((value1 + value2)/2), w isthe normalized block size in terms of a reference fre-quency ((log2(frequency/reference)),x is the normal-ized hue contrast ((|hue1 hue2|)/100),y is the valuecontrast (|value1 value2|), z is the average chroma((chroma1 + chroma2)/2), a = (y z), b = (v/5 + z/2),

and c = 1.3w.

Hue contrast (x), value contrast and average chroma

(a), and block size (c) account for the most rapid tran-sitions in the model.

The dimensions of color experience can be visualizedas spanning the multidimensional space described byf(v,w,x,y,z) in which all experiences are represented.Experiences comprise a response space that indi-cates the magnitude of the responses for a given com-bination of dimensions. The space spanned by thedimensions of experience provides the framework forestablishing any color experience because it indicateshow experiences relate to one another. Establishingcolor experience amounts to a navigation in thespace. For instance, traversing the space along a

region of constant responses results in experiencesthat are equivalent to one another. Navigation is per-formed by adjusting the contributions of each dimen-sion to the overall experience, and varying theposition in the space.

A plot of Equation 1 is shown in Figure 14. Examplesof rules for adjusting relative position in the space areshown in Figure 15. Figures 16 and 17 illustrateexample applications of the model.

Figure 16 The same reference hue, pink in thisexample, can carry two very different

messages: (A) baby colors are a result ofdecreasing hue and value contrast, and (B)soccer colors are a result of increasing hueand value contrast.

Figure 17 There are multiple means to affectexperience: changing (A) chromaticattributes, (B) spatial attributes, or (C) both.

(A) BABY COLORS (B) SOCCER COLORS

(A) (B) (C)

ORIGINALSTIMULUS


12/13


An intelligent tool for selecting colors

Color, when once reduced to certain definite rules, can betaught like music.

G. Seurat

People who think and create on an abstract level tendto experience difficulties when they cannot readilytranslate their abstract thoughts into tangible details,e.g., the nonartists dilemma in attempting to draw asad, frustrated, or understanding face. Most of us donot have the skills necessary to translate these simpleabstract notions into a concrete drawing.

Most people are color-inarticulate as well. Few of usare able to create consistent and robust messages withcolor. However, through the use of an expressivemodel, within a system of constraints, a person mightimprove his or her ability to compose a visual mes-sage that conveys his or her abstract thoughts.

The rules that govern color appearance, context, andexpression can be used as the basis for interactivegrammars. These grammars can be used to performhigh-level tasks, which are the foundation for commu-nicating with color messages, such as detection, legi-bility, and categorization, as well as expression.

One example of an intelligent tool for selecting colorsis the Color Rule and Font Tool (CRAFT),23-25 which isbeing developed by Rogowitz et al. as a general archi-tecture for adding guidance to interactive systems thathave been extended to the domain of user interfacedesign. The CRAFT system constrains choices for eachoperation (e.g., selection of background color and textcolor) based upon perceptual26 rules. CRAFT sharesmany design goals with other automated or guideddesign systems, such as those described by Mackin-lay,27 Feiner and McKeown,28 Roth and Mattis,29

Weitzman and Wittenberg,30 and Ishizaki.31 CRAFTsparticular emphasis is on the legibility and aestheticsof typography.

In the CRAFT architecture, all operations are linked;

every time the user makes a selection, the impact ofthat choice is reflected in the selections for other oper-ations. For example, if the user selects a color for awindow background, that information is fed back tothe operation responsible for selecting text color,where legibility rules constrain choices to ensure suf-ficient luminance contrast for good legibility. Addingrules for expression to the CRAFT architecture wouldbe a way to ensure that the visual messages generatedfrom an interactive system are effective.

Conclusion

Color demands a response. N. Jacobson

A future direction for this work is a study of how theexperience of color is related to emotion. The dimen-sions of color experience constitute the internal con-text in which a message is received. Burling andBender32 argue that this context establishes a sense-of-order or expectation. The violation of expectationand the subsequent resolution of discrepancy is anemotion mechanism of the type described by Mand-ler.33

Acknowledgments

This work was sponsored in part by InternationalBusiness Machines Corporation. The authors thankUri Feldman and Bil Burling for their contributions tothis work. The authors thank Bernice Rogowitz forher encouragement and advice.

**Trademark or registered trademark of Pantone or Macbeth.

Cited references

1. O. Sacks,An Anthropologist on Mars: Seven Paradoxical Tales,Knopf, New York (1995).

2. R. Evans,An Introduction to Color, John Wiley & Sons, Inc.,

New York (1948).3. S. Zeki, A Vision of the Brain, Blackwell Scientific Publica-

tions, Oxford (1993).4. D. H. Hubel,Eye, Brain and Vision, W. H. Freeman, New York

(1988).5. Munsell Book of Colors, Munsell Color Co., Baltimore, MD

(1942).6. G. Wyszecki and W. S. Stiles, Color Science: Concepts and

Methods, Quantitative Data and Formulae, Wiley IntersciencePublishers, New York (1982).

7. A. Hrd and L. Sivik, NCSNatural Color System: A Swed-ish Standard for Color Notation, COLOR Research and Appli-cations 6, No. 3, 129138, John Wiley & Sons, Inc., New York(1982).

8. E. Adelson, Perceptual Organization and the Judgment ofBrightness, Science 262, 20422044 (1993).

9. E. H. Land, The Retinex Theory of Color Vision, Scientific

American 237, No. 6, 108128 (1977).10. J. Y. Lettvin, P. Brou, T. R. Sciascia, and L. Linden, The Col-

ors of Things, Scientific American 246, 8491 (1986).11. R. Hunt, A Model of Colour Vision for Predicting Colour

Appearance, COLOR Research and Applications 7, No. 2, 95112, John Wiley & Sons, Inc., New York (1982).

12. B. A. Wandell, Foundations of Vision, Sinauer Associates, Sun-derland, MA (1995).

13. N. Jacobson,Relative Harmony Quotient, unpublished manu-script (1985).

14. S. Kobayashi, Color Image Scale, Kodansha, New York(1990).


13/13


15. P. Moon and D. Spencer, Geometric Formulation of ClassicalColor Harmony,Journal of the Optical Society of America 34,No. 242, 4659 (April 1944).

16. L. Sivik and A. Hrd, Semantic Variables for Judging ColorCombinationsAn Analysis of Semantic Dimensions, Gte-borg Psychological Reports 19, No. 5, University of Gteborg,Sweden (1989).

17. N. Jacobson, W. Bender, and U. Feldman, Alignment andAmplification as Determinants of Expressive Color, Proceed-ings of SPIE: Human Vision, Visual Processing, and Digital

Display, San Jose, CA, 1453 (1991).18. U. Feldman, N. Jacobson, and W. Bender, Quantifying the

Experience of Color, Proceedings of SPIE: Human Vision,Visual Processing, and Digital Display, San Jose, CA, 1913,537547 (February 1993).

19. P. Green-Armytage, A Game to Introduce the Language ofColour,Inter Society Color Council Annual Meeting, Prince-ton, NJ (1992).

20. C. E. Osgood, G. J Suci, and P. H. Tannenbaum, The Measure-ment of Meaning, University of Illinois Press, Urbana, IL(1957).

21. N. Jacobson, W. Bender, and B. Burling, Beyond ColorAppearance, Proceedings of SPIE: Human Vision, Visual Pro-cessing, and Digital Display, San Jose, CA, 2411, 6473 (Feb-ruary 1995).

22. J. J. Chang and J. D. Carroll, Three Are Not Enough: AnINDSCAL Analysis Suggesting that Color Space Has Seven

(1) Dimensions, COLOR Research and Applications 5, No.4, 193206 (Winter 1980).

23. B. E. Rogowitz and L. A. Treinish, An Architecture for Rule-based Visualization, Proceedings of the IEEE Conference onVisualization 93, San Jose, CA (October 1993), pp. 236243.

24. B. E. Rogowitz and D. A. Rabenhorst, CRAFT: A Tool forCustomizing Color and Font Selection Guided by PerceptualRules, Proceedings of IS&T and SID Color Imaging Confer-ence, Scottsdale, AZ (1993).

25. B. E. Rogowitz, J. A. Gerth, and D. A. Rabenhorst, An Intelli-gent Tool for Selecting Colors and Fonts for Graphical UserInterfaces, Proceedngs of SPIE: Human Vision, Visual Pro-cessing, and Digital Display, San Jose, CA, 2411, 3543 (Feb-ruary 1995).

26. B. E. Rogowitz, The Human Visual System: A Guide for theDisplay Technologist, Proceedngs of SID 24, No. 3, 235252(1983).

27. J. Mackinlay, Automating the Design of Graphical Presenta-tions of Relational Information,ACM Transactions on Graph-ics 5, No. 2, 110141 (1986).

28. S. K. Feiner and K. R. McKeown, Automating the Generationof Coordinated Multimedia Explanations, Computer24, No.10, 3341 (1991).

29. S. F. Roth and J. Mattis, Automating the Presentation of Infor-mation, Proceedings of the Fourth International Conferenceon Expert Systems for Production and Operations Manage-

ment, Miami Beach, FL (February 1991), pp. 2428.30. L. Weitzman and K. Wittenberg, Relational Grammars for

Interactive Design, IEEE Symposium on Visual Languages,Bergen, Norway (1993).

31. S. Ishizaki, Multiagent Model of Dynamic Design: Visualiza-tion as an Emergent Behavior of Active Design Agents, toappear inACM SIGCHI(1996).

32. B. Burling and W. Bender, Violating Expectations of ColorOrder, Proceedings of SPIE: Human Vision, Visual Process-ing, and Digital Display 2657, San Jose, CA (1996).

33. G. Mandler, Mind and Body: Psychology of Emotion andStress, Norton, New York (1984).

Accepted for publication May 13, 1996.

Nathaniel Jacobson (Mr. Jacobson died during the preparationof this paper.) Mr. Jacobson, an artist, teacher, and colorist, wasfounder and director of The Arts Students Workshop in Boston formore than twenty years. His work included paintings and murals invarious media, as well as designs for mosaics and stained glass. Hispaintings have received honors in national exhibitions and havebeen featured in one-man shows internationally. He lectured widelyto academic and industrial audiences. Mr. Jacobson maintained adeep interest in the study of color since his student days at Massa-chusetts School of Art, where he worked under the guidance ofAnne Hathaway, a disciple of Albert H. Munsell, and graduated in1937. He received the B.F.A. degree from Yale University in 1941.He was author ofThe Sense of Color(Van Nostrand Reinhold Com-pany, New York, 1975). As a color consultant since 1970 for Bin-ney & Smith Inc., manufacturers of art education materials, Mr.Jacobson guided the development of a special color range of artistspaints and color maps based on his patented Modular Color system.

He began working with computer color at the Architecture MachineGroup at MIT in 1980 and was a research affiliate at the MIT MediaLaboratory.

Walter Bender MIT Media Laboratory, 20 Ames Street, Cam-bridge, Massachusetts 02139-4307 (electronic mail: [email protected]). Mr. Bender is a principal research scientist at theMIT Media Laboratory and principal investigator of the labora-torys News in the Future consortium. He received the B.A. degreefrom Harvard University in 1977 and joined the ArchitectureMachine Group at MIT in 1978. He received the M.S. degree fromMIT in 1980. Mr. Bender is a founding member of the Media Labo-ratory.

Reprint Order No. G321-5622.

Documents

Color as a Determined Communication