The RC ColorChecker Calculator


 

The calculator is intended to be used to compare spot readings from a printed target to reference values. The latter may be published values or user-measured. Lab values from a digital image may also be compared. Color differences are calculated using either the CIE 2000 or 1976 equations. Graphic displays include the a-b difference, and the measured colors both as-is and L-corrected. The latter may be viewed in either sRGB or Display P3; all the patches should be in-gamut in the latter.

Begin by entering the color data. The default reference values are X-Rite’s published values for post-November 2014 targets. Measurements may be taken with any spectrophotometer supported by ArgyllCMS; the command is:

$ spotread logfile.txt

(If your instrument uses illuminant M0, M1, or M2, add the -I flag:)

$ spotread -I M2 logfile.txt

After calibrating if necessary, measurements are taken:

Place instrument on spot to be measured
. . .
Hit ESC or Q to exit, instrument switch or any other key to take a reading: 

 Result is XYZ: 9.433963 8.322203 4.247213, D50 Lab: 34.645098 12.100103 12.915079

Start with patch 1 (the brown one) and read each patch in the proper order (typically in rows, but columns work too). After exiting spotread, the measurements are written to the log file:

Reading	X	Y	Z	L*	a*	b*
1	9.433963	8.322203	4.247213	34.645098	12.100103	12.915079
2	31.709343	28.078042	15.207202	59.959179	17.714633	17.137164
3	12.776841	13.980994	18.801245	44.205662	-4.596293	-18.365140
. . .

Multiple sets of 24 readings, for both the reference and the measured target, will be averaged by the script; the measured values below consist of three sets.

Lab values from a digital image of the target may also be used. These may be either the raw values, or scaled values such as the 16-bit (PhotoLine) or 15-bit (Photoshop) presets below. Format the data as white space delimited L, a, and b values, one set per line; select which columns in the data to read; and enter scaled ranges as necessary. The decimal separator may be either dot or comma.

Please see more documentation below.


Enter reference and measured values:

reference values:  default

format:
start with column
  15-bit  16-bit
raw values scaled values
L: - -
a: - -
b: - -

measured values:

format:
start with column
  15-bit  16-bit
raw values scaled values
L: - -
a: - -
b: - -


Calculate the color differences:

calculate    |  sort by: 

R (reference) or M (measured) = L, a, b,
                                C, H



The a-b plane:




The actual colors:

  (will only be accurate in a Display-P3-enabled browser and a wide-gamut display)

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
reference measured measured, L-corrected

a b Δab ΔC ΔH
The diagram is a projection onto the a-b plane.
ΔL is the distance between the points on the z-axis.
ΔE is the three-dimensional distance between the points.
ΔE₀₀ is adjusted to be more perceptually uniform.

The color difference table data begins with the averaged reference and measured L, a, b, Chroma, and Hue values. ΔE is displayed as the native three-dimensional value and the a-b component. ΔL, ΔC, and ΔH are the Lightness, Chroma, and Hue differences, and Δab is the linear a-b distance.

The ΔE a-b component is the same as Δab in CIE 1976; in CIE 2000, it is the ΔE₀₀ equation without the ΔL′ term (ΔL′ = 0).

The table may be sorted by the Lightness, Chroma, or Hue of the reference colors, or by any of the Δ values. Note that patch 22 is the most neutral patch, and may serve as a “gray card.” The grays are not included when sorting by Hue.

The a-b plane includes the outline of the RGB color space (either sRGB or Display P3) at the L value of patch 18 (approximately 50); the patch should be in-gamut in the latter. The reference squares are the averaged values, if multiple are provided. A circle is drawn for every measurement to visualize the scatter. The radius of the circles is equivalent to Δab = 1. The ellipses are the ΔE a-b component at 1 and 2 (CIE 2000) or 2.3 and 4.6 (CIE 1976); these are intended to represent 1× and 2× the “just noticeable difference.”

The actual-color graphic consists of the averaged measured colors superimposed on the reference patches. The actual color is on the left; the color on the right, maybe hard to see, consists of the measured a and b values with the L value of the reference patch.

As of November 2014, X-Rite has issued new color specifications for ColorChecker targets. Unfortunately, much of the existing literature uses the old values. The default reference values above are the new ones; note that they are ordered in columns. (The script determines whether the data are read in rows or columns by the sign of a in patch 7.) If you have a closed-loop workflow, you can use your own measurements of the target.

Spot readings of printed targets are not supported by all device software, but the open-source ArgyllCMS comes to the rescue (be sure to support it!) and may be used with most hardware.

There are other color calculators, such as Imatest (Colorcheck | reference) which performs detailed analysis of digital images and camera performance. See also Danny Pascale’s BabelColor ColorChecker pages. If you have a need for the RGB values of the patches, download the X-Rite 16-bit Lab image, open it in an image editor such as PhotoLine, and convert it to the desired working space.

The CIE 2000 equations are taken directly from Gaurav Sharma, Wencheng Wu, Edul N. Dalal, “The CIEDE2000 Color-Difference Formula: Implementation Notes, Supplementary Test Data, and Mathematical Observations,” Col Res Appl, 30, 21–30, 2005. See Dr. Sharma’s page for a link and more information.



More references to the equations and what they are:

Delta E (CIE 1976) and Delta E (CIE 2000) at brucelindbloom.com

CIE76, CIEDE2000, and CIELCh at Wikipedia

A few Amazon affiliate links:

ColorChecker Classic Mini

the Calibrite store

the X-Rite store


russellcottrell.com/photo