Maps Ch 9: Color Theory

what are the two main categories of color theory?

traditional color "theory" and scientific "color theory"

diverging schemes

used for bipolar data

sequential schemes

used for unipolar data

scientific "color theory"

various scientific disciplines investigating color (color vision, psychology of color, psychophysics, colorimetry, etc.)

types of color models

- device oriented - perception oriented

color models

- "tools" used to specify color - models are given coordinate systems in color space: > they give colors numerical coordinates that correspond to brightness, saturation, etc. and can be adjusted for precise color selection > each color model has unique properties that make it useful in a particular way

types of color models: perception oriented

- Munsell - CIELab - CIELuv

types of color models: device oriented

- RGB, CMYK - CIEXYZ - CIExyY

advanced color considerations: color defieciencies

- anomalous trichromats: some difficulty distinguishing colors (esp. red and green) protanomalous and deuteranomalous groups - anomalous dichromates: complete inability to distinguish red and green, among other color difficulties

color schemes fo specific tasks

- certain data types require special consideration when picking a color scheme: > phenomena that have "natural" colors associated with them > phenomena shown with "traditional" colors > color for "no data" values

Traditional color "theory"

- collection of advice about mixing and the effects of resulting color combinations (1700s and afterwards) - centered around two key concepts: > primary colors > color wheel - strong artistic tradition: > traces back to da Vinci (1400s) > developed by Newton et al. (1700s) - limited to how color really works!

CIELuv/Lab properties

- communicate the exact values of color, with all its complexity - aligns luminance with the depth axis: > makes it convenient to get "darker" and "lighter" colors - both are perceptually even: adjacent colos appear similar, easy to make perceptually smooth color schemes *not tru of CIEXYZ or xyY

color scheme design for data mapping

- key consideration is matching the nature of the phenomena to the nature of color scheme: > sequential schemes -- unipolar data > diverging schemes -- bipolar data > qualitative schemes -- categorical data - perceptual issues to consider: > equal steps in "color" for equal steps in data > clear separation of opposite ends of scale > color deficiencies, simultaneous contrast etc.

types of color models: device oriented: RGB

- mirrors the hardware of a typical screen (makes pixel manipulation trivial) - has significant limitations: > no natural relation to hue, brightness or saturation > does not capture all visible colors > perceptually UNEVEN

additive color theory

- mixing light

subtractive color theory

- mixing of ink example: CMY(K) primary colors

CIE family

- original CIE color models: > CIEXYZ: first to define quantitative links between physical colors and percieved colors > CIExyY - perceptual CIE color modols: > CIELab and CIELuv for paper and screen, respectively

cultural conditioning and color vision

- people are born with certain capacity for color vision, but it changes as you learn to speak - language "color scheme" will define your capacity for color vision (culture specific) - cognitive linguistics studies this - color psychology investigates universal color connotations, but they are very subtle

issues with traditional color theory

- selection of primary colors is driven by technological limitations: > availability of pigments (inks) to print and dye > availability of phosphors and LEDs for screens - as technology improves, primaries change: > RYB to CMYK > RGB to RGBYC - color vision is a complex process: > studies of eye physiology showed we perceive differences between colors rather than colors themselves

color scheme selection tools

- sequential, diverging and qualitative schemes: > color brewer - sequential: > NASA Ames Color Tool > Chroma.js > Colorpicker for data" - qualitative: > I Want Hue

types of color models: perception oriented: munsell model

- uses hue, value, and chroma as coordinates: > "color", "brightness" and saturated - perceptual model designed before computer age: > adjacent colors are perceptually equidistant > asymmetric due to quirks of our color vision

advanced color considerations: simultaneous contrast

- when one color is surrounded by a different color: > brightness shifts away from the dominant neighbor > hue shifts toward complementary of the dominant neighbor aka induction

advanced color considerations

1. color deficiencies and how to deal with them 2. simultaneous contrast 3. color illusion

what is CIE?

CIE is the Intl. Commission on Illumination (Commission International de l'Eclairaige)

qualitative schemes

categorical data

Helmholtz-Kohlrausch effect

intense saturation is perceived as part of the sample's brightness ex: don't use rainbows

alternative colors and wheels

many primary colors possible: - RYB (red, yellow, blue) - RGV (red, green, violet): young - RGB (red, green, blue): maxwell multiple ways of mixing primary colors: - subtractive and additive, the difference is mixing ink vs light

concepts based on the color wheel

primary, secondary and tertiary colors: - the latter two are achieved by progressive mixing of neighboring colors analogous and complementary colors: - adjacent and opposite on color wheel color scheme recommendations: - primarily based on analogous and complementary concepts

primary colors

supposedly, any color can be produced by mixing a proportion of primary colors together (this should somehow hold true for paper, screens, etc)