Should I add a color calibration target to some figures in a scientific paper?

Question

I have to say first that I am not very familiar with the world of color rendering in computer graphics, I am more closer to a physical approach of the phenomenon.

My problem is that I am writing a scientific paper in which we are recreating the perceived colours of some materials, and in which of course we are showing our results in some figures... But as always in this kind of things, how can I be sure that the colours of the figure that I am creating on my laptop will look the same on the screen of readers of my article, or on printed papers?

I know that this is impossible, but my idea was to add next to the figure a color calibration, such as a simple black-yellow-magenta-cyan square, or a ColorChecker figure... I do not know what is the convention for this kind of problems, or if one of my solutions is better than the other.

If someone has an advice, I'll take it!

ColorChecker Example

CMYK Example

Answer 1

With the possible exception of specific fields, I think you should expect most readers to read the paper on non-professional displays that lack color-calibration or full bit depth, or on a color-inaccurate printout. I've also seen publishers end up affecting color balances when compressing figures for publication in the form of .pdf files. The case of actual print journals is rare enough these days that you can probably neglect it, but if it applies you should ask them how to prepare figures for accurate color reproduction with their printing processes.

If the precise color information really matters for your paper, I think including a calibration target in generated graphics or a color rendition chart in photographs would be reasonable. Whether you include one or not, however, it would also seem worth supplying the figures either in the form of supplemental material deposited with the paper, or in a data archive cited in the paper. (If you go the data archive route, you may need to clear it with the publisher beforehand depending on how the figures are copyrighted.) That is a more robust way of communicating the exact color information, and more convenient to readers that would make use of it. Alternative ways of communicating the colors, such as including color codes in an appendix, might be workable too.

Of course, in the majority of cases good enough is... good enough. Certainly for plots and illustrations you should design them such that they're easy to read (also for color blind readers!) and don't require ideal color reproduction.

Answer 2

In general the advice is: Assume people print out your paper in black-and-white. They should not lose any information.

However it seems your specific paper concerns color in a direct way so this doesn't really apply. Here I would recommend to specify the exact colors you mean through one of the standardized notations like RGB or Hex or something like that. RGB (123, 23, 43) is a precise specific color. A given computer screen or printer might not render it correctly but that is on the device not on you. It is uniquely specified which color you mean.