In a previous post we discussed the use of the X-Rite ColorChecker SG card and the online platform delt.ae as a means to control and validate the colour accuracy of our digitisation workflows. Whilst that process allows for a quick colour management, we found it to be somewhat insufficient in those cases where we want to take the quality of our captures to the next level.
The problem lies in the data that is used to construct and validate the ICC colour profiles. These profiles are used to map the colours that are produced by the camera against the colours that the camera should produce and are constructed using BasICColor Input. This software comes with a whole array of generic value sets for different colour targets, including the X-rite ColorChecker SG card. Convenient, one would think. And it is: using the supplied value sets, it is possible to pass the highest Metamorfoze validation level on delt.ae.
However, here at the Imaging Lab we are not satisfied with simply passing a guideline. We aim to achieve the best results possible. Therefore, basing the construction of our ICC profiles on measurements made by someone in a far-off lab under unknown circumstances just seems inadequate. Moreover, colour targets fade due to age and individual targets can show some deviations as a result of changes or inconsistencies in the production proces.
The solution is to take measurements into our own hands. Using an X-Rite EyeOne spectrophotometer and SpectraShop 5 software, we were able to measure our own Colorchecker cards. Using our own measurements as a base for our ICC profiles, we can achieve (even more) accurate and faithful reproductions.
Below you will find three screenshots that demonstrate the impact of a good colour management. The numbers show the ΔE values of the colour patches of the ColorChecker SG. ΔE expresses the difference between the actual colour of a patch and how it is reproduced by the camera. A lower number thus means a more accurate reproduction. The Metamorfoze guidelines specify a mean ΔE lower than 4 and a max ΔE of 10.