H517 Visualization Design, Analysis, & Evaluation Week 4: - PowerPoint PPT Presentation

H517 Visualization Design, Analysis, & Evaluation Week 4: Color Perception Khairi Reda | redak@iu.edu School of Informa5cs & Compu5ng, IUPUI

Paper presentations • Every week we will have 2 papers, each paper will be presented by two students • Prepare a PowerPoint presenta5on about the paper ( 15 minutes ) • Cri5que the authors’ work: What are the main claims? Did the authors present enough evidence to back them up? Do you like the paper? What would you change about it? • Lead a discussion (15 minute) • End your presenta5on with ini5al ques5ons for the class • Everyone should read the paper before coming to class • Asking ques5ons and providing comments will count towards your 10% par5cipa5on grade

Last week…

Visual perception light electricity http://www.fortworthastro.com/images/eye_xsection_01.jpg

C. Ware Visual Thinking for Design

POPOUT

POPOUT

Take home point Our visual system sees differences, not absolute values Based on a slide by Miriah Meyer

Visual perception is relative Differences in contrast is rela5ve

Visual perception is relative Sizes are rela5ve (Ames room)

color The Huffington Post

color the property possessed by an object of producing different sensa5ons on the eye as a result of the way the object reflects or emits light -Oxford dictionary

light Electromagne5c radia5on within a certain range [400nm - 700nm] of the electromagne5c spectrum more energy

light Mariah Meyer

Trichromacy

Trichromacy Normal human color vision is 3 dimensional Derived from three cone types (short, medium, and long wave-length sensi5vity) Each type of cone contains a specific photosensi5ve pigment that reacts to a certain wavelength of light Based on a slide by Mariah Meyer

Trichromacy easy to read difficult to read easy to read difficult to read

Opponent-process theory Explains how signals are processed Visual perceptual system detects differences in the response of cones + + - - red-green yellow-blue luminance opponent channel opponent channel

Opponent-process theory Explains how signals are processed Visual perceptual system detects differences in the response of cones + + - - red-green yellow-blue luminance opponent channel opponent channel

“Important” colors These colors have a name in virtually every human language Their seman5cs and connota5ons are culture- specific

Sensitivity to spatial detail The luminance channel has greater ability to resolve smaller detail C. Ware, Visual Thinking for Design

Sensitivity to spatial detail C. Ware, Visual Thinking for Design

Color deficiencies

Color deficiencies Some5mes caused by faculty cones, some5mes by faulty pathways red-green weakness is the most common type 8% of (North American) makes, 0.5% of female normal re5na Protanopic Based on a slide by Miriah Meyer

lacking red cones lacking green cones lacking blue cones Via Miriah Meyer

difficult to dis5nguish for people with Deuteranopia

Color spaces Representing color with numbers

light 1. pure yellow: 580 nm 2. color matching yellow

Tristimulus color matching red green test color blue

Tristimulus color matching red green test color 580nm blue

Tristimulus color matching red 0.17 green 0.17 test color 580nm blue 0

Tristimulus color matching red green test color 580nm blue

Tristimulus color matching red green test color 580nm blue

RGB color space Each point within the cube is defined by a 3D vector (r, g,b) Each point within the cube is defined by a 3D vector (r, g,b) and represents a unique color and represents a unique color The r, g, b coordinates of the vector reflect a combina5on of The r, g, b coordinates of the vector reflect a combina5on of red, green, and blue primaries needed to reproduce the color red, green, and blue primaries needed to reproduce the color 1.0 G 1.0 R B 1.0

RGB color space Each point within the cube is defined by a 3D vector (r, g,b) and represents a unique color The r, g, b coordinates of the vector reflect a combina5on of red, green, and blue primaries needed to reproduce the color yellow (1.0, 1.0, 0.0) 1.0 1.0 1.0

RGB color space Each point within the cube is defined by a 3D vector (r, g,b) and represents a unique color The r, g, b coordinates of the vector reflect a combina5on of red, green, and blue primaries needed to reproduce the color 1.0 (1.0, 0.6, 0.4) 1.0 1.0

RGB color space Each point within the cube is defined by a 3D vector (r, g,b) and represents a unique color The r, g, b coordinates of the vector reflect a combina5on of red, green, and blue primaries needed to reproduce the color 1.0 1.0 white (1.0, 1.0, 1.0) 1.0 1.0 1.0 1.0

RGB color space Each point within the cube is defined by a 3D vector (r, g,b) and represents a unique color The r, g, b coordinates of the vector reflect a combina5on of red, green, and blue primaries needed to reproduce the color 1.0 1.0 1.0 1.0 black (0.0, 0.0, 0.0) 1.0 1.0

what colors combinaCon can be used to re- producing the visible light spectrum by mixing? • red, yellow, blue • red, green, blue • orange, green, violet • cyan, magenta, yellow • all of the above Miriah Meyer

Light mixing (RGB) Addi5ve mixing of colored lights

Light mixing (RGB) LCD display closeup Wikipedia

Ink mixing (CMY / CMYK) Subtrac5ve mixing of inks printed on white paper

CMY composite CMYK composite Color picture Wikipedia

what colors combinaCon can be used to re- producing the visible light spectrum by mixing? • red, yellow, blue • red, green, blue • orange, green, violet • cyan, magenta, yellow • all of the above , almost Miriah Meyer

Tristimulus color matching red green test color 500nm blue

Tristimulus color matching red green test color 500nm blue

Tristimulus color matching red green test color 500nm blue

Tristimulus color matching red green test color 500nm blue

Opps

CIE color space • At a mee5ng in of the CIE in 1931 • Let’s have imaginary primary colors! • Construct linear, possibly non-realizable combina5ons of primaries so that color matching func5ons are posi5ve throughout the visible light • X, Y, Z primaries • Can be linearly transformed from RGB (and vice versa) Based on a slide by Siddhartha Chaudhuri

CIE color space

1.0 Y Y 1.0 X X Z Z 1.0

Y X Z

CIE chromaticity diagram

CIE chromaticity diagram White

CIE chromaticity diagram White

CIE chromaticity diagram G Y R

CIE chromaticity diagram RGB color space

Perceptual color spaces A change in the amount of color value should produce a propor5onal change in the way we see the color Via Miriah Meyer

HSL • hue: what people think of as color • saturation: the vividness of the color • luminance: amount of black mixed in

hue saturation luminance

Guidelines for using color in visualization

Colormap Specifies a mapping between [0, 8] color and values

Order these colors… Miriah Meyer

Order these colors… Miriah Meyer

Order these colors… Miriah Meyer

guidelines colormaps for ordered data should vary monotonically in luminance Hue alone is good for categorical data Categorical colors are easier to remember if they are nameable Colin Ware

the rainbow colormap temperature

the rainbow colormap order?

the rainbow colormap sharp boundary

the rainbow colormap not color blind safe

the rainbow colormap Rainbow colormaps should be avoided as a default op5on for ordered data A safer, more effec5ve op5on is a colormap that varies in luminance. Ideally luminance and hue .

Colin Ware

Simultaneous contrast Via Colin Ware

Color design tools Colorgorical http://vrl.cs.brown.edu/color Color Brewer colorbrewer2.org/ Color Oracle http://colororacle.org

Next week Marks and channels: the visualiza5on alphabet (chapter 5)

D3 lab hUps://Cnyurl.com/y9z8gefr hUp://vis.ninja/teaching/2018/H517/d3-excercises/