The performance of color matching in ICC-based Color Management Systems (CMS) is being improved, and its workflow established. As part of the digital workflow, digital proofing plays an important roll in the control of color from capture to the printed page, but its method of evaluation is not clearly understood.

To evaluate the degree of color matching by digital proofing in an ICC-based CMS between a reference image and a sample image, the [formula] calculated by averaging the individual color differences is being used. Doing a qualitative evaluation, ICC-based CMS should be able to stand behind the statement that the color space transformation through the Profile Connection Space (PCS) is correlated to visual judgments. The ΔE*a,b Distribution should provide more information than the average [formula] to evaluate the degree of color matching in digital proofing.

In this research, the visual judgment between digital proof and a reference image color matches is studied. Using a reference image, and digital proofs as sample images, a psychometric evaluation based on the one-dimensional scaling technique is applied to scale the visual assessments. From the three dimensions: lightness, chroma, and hue; the lightness (L*) dimension was the only aspect addressed to scale the visual assessments.

The reference image was generated with the color characterization data for Type 1 printing 1, and digital proofs with the input data characterization of 4-color process printing (IT8.7/3 target)2,3. Those digital proofs were presented to a panel of observers; data were collected, and analyzed to develop a preference scale that represents the qualitative evaluation of the color match.

To test the accuracy of the ΔE*a,b Distribution for evaluating the degree of color matching, a quantitative analysis was done. The ΔE*a,b Distribution was analyzed by the cumulative frequency distribution technique between the reference image and the digital proofs, for 182 patches of the IT8.7/3 target, and the best critical value to match the visual scale was obtained. A quantitative evaluation between the digital proofs and the reference image, based on the color characterization data for Type 1 printing 1,4, was used to scale the average [formula] and the ΔE*a,b Distribution. Then, a classical statistical method was applied to those metric scales and to the visual scale to infer that the degree of color matching is better correlated between the visual judgement and the ΔE*a,b Distribution, than between the visual judgement and the average [formula].

Finally, using this method based on the qualitative analysis, a median of the ΔE*a,b Distribution of 52.8% would produce a visual match between the reference and the sample.

Endnotes for Abstract

  1. ANSI CGATS TR 001-1995 Graphic technology - Color characterization data for Type 1 printing.
  2. ISO 12642:1997 Graphic technology - Prepress digital data exchange - Input data for characterization of 4-colour process printing.
  3. ANSI IT8.7/3. Graphic technology - Input data for characterization of 4-color process printing. Reston, Virginia. June 21 (1993).
  4. ANSI CGATS.6-1995 Graphic technology - Specifications for graphic arts printing Type 1.

Library of Congress Subject Headings

Proofs (Printing)--Digital techniques; Color printing--Digital techniques; Electronics in color printing; Image processing--Digital techniques

Publication Date


Document Type


Student Type



Robert Y. Chung


Alt Title: Using DE Distribution as a predictor of digital proofing performance

Physical copy available from RIT's Wallace Library at Z258 .D67 2000


RIT – Main Campus