Perceptual Colormaps

How science got off jet. Borland and Taylor diagnosed the problem, Dave Green's cubehelix and Moreland's cool-warm offered early fixes, and the SciPy 2015 viridis talk made the argument impossible to…

Summary

How science got off jet. Borland and Taylor diagnosed the problem, Dave Green's cubehelix and Moreland's cool-warm offered early fixes, and the SciPy 2015 viridis talk made the argument impossible to ignore. Kovesi, Crameri, and Thyng then built whole families whose lightness actually tracks the data, and Turbo is the honest rainbow for the cases where you still want one. Read the talk first, then pick a family and install it.

Entries

  1. A Better Default Colormap for Matplotlib: The SciPy 2015 talk: the case against jet, and how viridis was designed in CAM02-UCS. (source)
  2. A Better Default Colormap for Matplotlib: The case for viridis: why jet misleads, and the perceptual constraints on a good default. (source)
  3. How Bad Is Your Colormap?: A short function that renders any Matplotlib colormap in grayscale, exposing jet’s bands and reversals. (source)
  4. Rainbow Color Map (Still) Considered Harmful: The classic case against rainbow maps: no perceptual order, hidden structure, and invented edges. (source)
  5. How NOT to Lie with Visualization: The rainbow draws contours the data lacks; match the colormap to the task and the data’s spatial frequency. (source)
  6. The Rainbow Is Dead… Long Live the Rainbow!: A series for geoscientists on what goes wrong in the rainbow, and how to build a perceptual one in CIELAB. (source)
  7. Cubehelix: A colormap that spirals around the grey axis while brightness rises steadily, so it survives black-and-white printing. (source)
  8. Diverging Color Maps for Scientific Visualization: The paper behind the cool-warm map: interpolate diverging scales in Msh so the midpoint stays neutral. (source)
  9. Good Colour Maps: How to Design Them: The paper behind ColorCET: a test image that exposes flat spots in a colormap, and a recipe for uniform ones. (source)
  10. ColorCET: The theory and full set of Peter Kovesi’s perceptually uniform colormaps, with a test image for each. (source)
  11. The Good, the Bad, and the Ugly: A Theoretical Framework for the Assessment of Continuous Colormaps: Turns colormap rules of thumb — order, uniformity, discriminative power — into definitions you can test. (source)
  12. Somewhere Over the Rainbow: An Empirical Assessment of Quantitative Colormaps: Nine colormaps tested on distance judgments: viridis holds up at every span; single-hue blues blur small gaps. (source)
  13. Mapping Color to Meaning in Colormap Data Visualizations: Viewers read darker as more — unless a map looks semi-transparent, and then more opaque reads as more. (source)
  14. Scientific Colour Maps: Perceptually uniform, colorblind-friendly maps such as batlow and vik, built to replace rainbow scales. (source)
  15. The Misuse of Colour in Science Communication: How rainbow and red–green maps distort scientific figures, and what to use instead. (source)
  16. True Colors of Oceanography: The paper introducing cmocean: one colormap per ocean variable, chosen to match how each quantity behaves. (source)
  17. cmocean: Oceanography colormaps — thermal, haline, dense, and more — each matched to the variable it encodes. (source)
  18. Optimizing Colormaps for Color Vision Deficiency: The paper behind cividis: a viridis variant tuned so color-vision deficiency doesn’t change how the scale reads. (source)
  19. Turbo, An Improved Rainbow Colormap for Visualization: A rainbow for when you need maximum hue contrast, with jet’s banding and false edges smoothed out. (source)
  20. Rainbows Revisited: Modeling Effective Colormap Design for Graphical Inference: Colormaps that pass through more nameable colors, rainbows included, made people more accurate at inference. (source)
  21. viscm: Design a colormap while watching its lightness curve and CVD simulations — the tool used to build viridis. (source)
  22. CCC-Tool: Build a continuous colormap from a specification, then test it with ΔE plots and test fields before export. (source)
  23. cmcrameri: Registers Fabio Crameri’s Scientific colour maps with Matplotlib, so batlow is one import away. (source)
  24. cmap: Colormaps from Matplotlib, cmocean, ColorBrewer, Crameri, and more, with NumPy as the only dependency. (source)
  25. viridis for R: viridis, magma, inferno, plasma, cividis, and friends as R palettes with ready-made ggplot2 scales. (source)
  26. ColorSchemes.jl: Julia’s colormap catalog — scientific, cmocean, Crameri, ColorBrewer, and more — with a preview of every ramp. (source)
  27. Palettize: A CLI and Python library for colormaps: terminal previews, CVD and perceptual checks, and export to GDAL and QGIS. (source)