Analysis of Error in Corrective Computerized Color Matching

Abstract

It is generally known that you need at least one additional CCM correction to match a color so as to get the final sample degree of accuracy below ΔE0.3 when the first matching accuracy is around ΔE1, and at least two additional corrections when the first matching accuracy is around ΔE2. However, there has been no specific report on the accuracy of CCM matching from the theoretical standpoint until recently. In this report, we tried to set out the specific behavior of CCM corrections theoretically under ordinary conditions when the abnormal coloring factors are eliminated from the related parameters; Pursuing more accurate CCM corrections there are five major factors that affect the CCM correction. The first is the calculation error stemming from the algorithm for the CCM corrective calculation (factor-A). The second is the error stemming from the deviation between the figures for the Kubelka-Munk theory and those for the actual ones (factor-B). The third is the error caused by production lots of the colorants (factor-C). The fourth is the error caused by coloring reproducibility (factor-D). And the fifth and final one is the error caused by the difference of the surface gloss of the matched samples (factor-E). As a result of our study, we have come to the conclusion that the errors caused by factors-A, B and C are negligibly small and the errors caused by factors-D and E would have direct effects on the degree of accuracy of CCM corrections. It is confirmed that provided you use colorants having good coloring reproducibility and you consider the reproduction of the surface gloss, it may be possible to obtain a final sample degree of ΔE0.3 with one correction even under the condition of the first matching accuracy being around ΔE2.