II-1Color in Scientific VisualizationMike [email protected] State Universitymjb – March 27, 2013Oregon State UniversityComputer Graphics“The often scant benefits derived from coloring dataindicate that even putting a good color in a goodplace is a complex matter. Indeed, so difficult andsubtlethatavoidingcatastrophebecomesthefirstsubtlethatavoidingcatastrophebecomesthefirstprinciple in bringing color to information.Above all, do no harm.”-- Edward Tuftemjb – March 27, 2013Oregon State UniversityComputer GraphicsII-2What’s Wrong with this Color Scale?mjb – March 27, 2013Oregon State UniversityComputer GraphicsSource:Scientific American, June 2000Not a bad choice of color scale,but the Dynamic Range needs some workmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-3Let’s start with the most important componentin a visualization system – You!How Many Shades of Different ColorsAre We Able to Detect?mjb – March 27, 2013Oregon State UniversityComputer GraphicsRods• ~115,000,000Sensors in Your Retina• Concentrated on the periphery of the retina• Sensitive to intensity• Most sensitive at 500 nm (~green)Cones•~7 000 000mjb – March 27, 2013Oregon State UniversityComputer Graphics•~7,000,000• Concentrated near the center of the retina• Sensitive to color• Three types of cones: long(~red), medium (~green),and short (~blue) wavelengthsII-4mjb – March 27, 2013Oregon State UniversityComputer Graphicsmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-5mjb – March 27, 2013Oregon State UniversityComputer Graphicsmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-6Sidebar: How Many Pixels Do You Need?A person with 20/20 vision has a visual acuity of:1 arc-minute = 1/60°36 9531 11124143Viewing Distance(inches)RequiredPixel Density(ppi)Density = D1Θ = 1/60° = .00029RIf the monitor’s resolution is 1600 x 1200, then its diagonal size would need to be:21”18”14”mjb – March 27, 2013Oregon State UniversityComputer Graphics2414312 286940066007”5”3”Monitors: Additive Colorsmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-7Additive Color (RGB)RC=G+BM=R+BW=R+G+BY=R+GGmjb – March 27, 2013Oregon State UniversityComputer GraphicsBglColor3f( r, g, b );0. ≤ r, g, b ≤ 1. OpenGL:Plasma Displays use Additive Color• Gas cell• Phosphormjb – March 27, 2013Oregon State UniversityComputer Graphics• Grid of electrodeshttp://electronics.howstuffworks.comII-8LCD Displays use Additive Color• Grid of electrodesmjb – March 27, 2013Oregon State UniversityComputer Graphics• Color filtershttp://electronics.howstuffworks.comDigital Film Recorders also use Additive ColorClick!Click!Cli k!Click!High-resolution FromComputerRBGmjb – March 27, 2013Oregon State UniversityComputer GraphicsClick!ggrayscale monitorColor wheelCameraII-9Hue-Saturation-Value:For many vis applications, a simpler way to specify additive color120ºHueWhiteSaturation240º1200ºWhitemjb – March 27, 2013Oregon State UniversityComputer GraphicsValueBlackfloat hsv[3], rgb[3];HsvRgb( hsv, rgb );glColor3fv( rgb );240The HsvRgb routine is in your sample code0. ≤ s, v, r, g, b ≤ 1.0. ≤ h ≤ 360.mjb – March 27, 2013Oregon State UniversityComputer GraphicsII-10120ºNotice that blue-green-red in HSV space corresponds to the visible portion of the lt ti tHue-Saturation-Value:For many vis applications, a simpler way to specify additive color240º0ºelectromagnetic spectrumBlue: 380 nm Green: 520 nm Red: 780 nmmjb – March 27, 2013Oregon State UniversityComputer Graphicsminmax min240. 240.SSHueSSTurning a scalar value into a hue when using the Rainbow Color ScaleHue-Saturation-Value:The ColorPicker Programmjb – March 27, 2013Oregon State UniversityComputer GraphicsRed, Green, Blue Hue, Saturation, ValueII-11The OpenDX Visualization Software Allows you toSculpt the Transfer Function in HSVmjb – March 27, 2013Oregon State UniversityComputer GraphicsSubtractive Colors (CMYK)BMCBGmjb – March 27, 2013Oregon State UniversityComputer GraphicsCBGRII-12Subtractive Color (CMYK)CR=M+YG=C+YK=C+M+YB=C+MMmjb – March 27, 2013Oregon State UniversityComputer GraphicsYColor Printing• Uses subtractive colors• Uses 3 (CMY) or 4 (CMYK) passesCMYK i ll h blki blk•CMYK printers usually have a better-looking black• There is a considerable variation in color gamut between productsmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-13How Do Color Separations Work in Color Printing?mjb – March 27, 2013Oregon State UniversityComputer GraphicsSource: R. Daniel Overheim and David Wagner, Light and Color, John Wiley & Sons, 1982.mjb – March 27, 2013Oregon State UniversityComputer GraphicsII-14mjb – March 27, 2013Oregon State UniversityComputer GraphicsGetting the CMYK ColorsWaxTTonerTonermjb – March 27, 2013Oregon State UniversityComputer GraphicsTonerSheetsII-15CIE Chromaticity Diagram0.800.90520 nmy0.100.200.300.400.500.600.70780 nmmjb – March 27, 2013Oregon State UniversityComputer Graphics0.00x0.00 0.20 0.40 0.60 0.80White Point380 nmCIE Chromaticity Diagram0.800.90520 nmy0.100.200.300.400.500.600.70780 nmCDC'mjb – March 27, 2013Oregon State UniversityComputer Graphics0.00x0.00 0.20 0.40 0.60 0.80White Point380 nmC = the colorD = the dominant wavelengthC ’ = the complementary colorII-16Color Gamut for a Workstation MonitorColor CRTWhite Point0700.800.90EyeMonitor Whitey0.100.200.300.400.500.600.70mjb – March 27, 2013Oregon State UniversityComputer GraphicsEye0.00x0.00 0.20 0.40 0.60 0.80Color Gamut for a Monitor and Color SlidesSlide White0800.90Color CRTProjected Color Slidesy0.200.300.400.500.600.700.80mjb – March 27, 2013Oregon State UniversityComputer GraphicsEyeWhite Point0.000.10x0.00 0.20 0.40 0.60 0.80II-17Color Gamut for a Monitor and Color Printer0800.90Color CRTColor Paper Hardcopyy0.200.300.400.500.600.700.80mjb – March 27, 2013Oregon State UniversityComputer GraphicsEye0.000.10x0.00 0.20 0.40 0.60 0.80The Perceptually Uniform L-a-b Color Space520 nmOSU Logo780 nmOSU Logomjb – March 27, 2013Oregon State UniversityComputer GraphicsWhite Point380 nmII-18Color Meters Are Able to Measure L-a-b Coordinatesmjb – March 27, 2013Oregon State UniversityComputer GraphicsSome Good Rules of ThumbWhen Using Color for Scientific Visualizationmjb – March 27, 2013Oregon State UniversityComputer GraphicsII-19What Makes a Good Contrast?• Many people think simply adding
View Full Document