node package manager



To access the latest version of d3-jnd, grab it on unpkg. Note that d3-jnd depends on d3-color.

This module extends D3 to support color just-noticeable difference (JND) research by Maureen Stone, Danielle Albers Szafir, and Vidya Setlur (link).

  title={An engineering model for color difference as a function of size},
  author={Stone, Maureen and Szafir, Danielle Albers and Setlur, Vidya},
  booktitle={Color and Imaging Conference},
  organization={Society for Imaging Science and Technology}

d3-jnd has two functions: d3.noticeablyDifferent() and d3.jndInterval. Both take size and percent arguments, which alter how conservative JND estimations are.

Note that the recommendations that are provided by this library are predicted JND intervals, and should be treated as guidelines rather than absolute truth. See the d3-jnd website for more information about how to manipulate d3-jnd function arguments appropriately.

var c1 = d3.lab('black'),
    c2 = d3.rgb('white'),
    legible = d3.noticeablyDifferent(c1, c2); // true
var intervals = d3.jndInterval(0.5, 0.3);
var strictlyLegible = d3.noticeablyDifferent(c1, c2, 0.95, 0.1);


After downloading the repo, run npm install, which will install any dependencies. You can optionally install from npm opposed to cloning directly from GitHub. Make sure to load d3-cam02 after d3-color.

Dependencies: d3-color

API Reference

# d3.noticeablyDifferent(color, color[, percent, size]) <>

A function that returns true/false based on whether the two provided colors are noticeably different. Percent controls the percent of audience that will likely see a difference between the two colors quickly and easily. Size refers to the estimated visual angle of color area (smaller areas are harder to discriminate).

# d3.jndInterval(percent, size) <>

As d3.noticeablyDifferent except it returns the intervals along L*, a*, and b* channels of CIELAB that produce differences equal to "1 JND".