Evaluation of digital proofs using the GRACoL dataset and various color difference equations
During the June 2007 IPA conference a proofing round up was conducted and it was verified that all manufacturers could match the colors of the GRACoL Coated 1 dataset based on measurement data only. The IT8.7/4 target consisting of 1617 different colors was measured with a DTP70 instrument, and color differences were determined with the Comparison Tool in X-Rite’s ProfileMaker software.
The test also included submissions from users, many of which also showed very similar, low differences. The criterion for evaluating the color differences was based on the DEab equation. The average DEab for all submitted proofs was 1.6.
A visual evaluation of the proofs submitted by vendors and users was also conducted. This evaluation showed that in some instances the measured data gave the same DEab, but visually the proofs were rated differently.
In this paper the data will be reevaluated using the color differencing equations of DE94, DE2000, and DECMC. It will be shown that it could be beneficial to use one of the aforementioned, newer color differencing equations as a criterion to rate the color accuracy of proofs. Work done in North America has focused on the use of DE94, DE2000 and DECMC, however more common in Europe is the work of the DIN and the related standards. In this paper the color differencing equation DIN99, which is specified in DIN 6176, and also in ASTM D2244-07, was used. This equation transforms the L*a*b*-color space to a uniform color space. It applies a warping to L*, a*, b* to get L99, a99, and b99 values. This new color space is more uniform so that color differences can be computed with the normal Euclidean distance formula.
The overall evaluation of the data showed that because of all the various criteria that had to be met using DEab metrics, these criteria are quite
