Some time ago, just starting out at a photographic company, I asked why essentially all media depictions of rainbows look fake. A color-theory expert there asked how saturated I thought rainbow colors might be. I guessed around 50%. He laughed, told me to sign up for the company's introduction-to-reprography seminar, and said that normal printing techniques would strain achieve that - needed Kodachrome, a good screen and a dark room to show anything so dramatic. He suggested that real rainbows only look colorful in context, with most of the apparent contrast emerging in brain rather than actual retinal signal; that even the most impressive rainbows, with dark-cloud backgrounds, rarely have a peak chroma/saturation above about 5%.

I have since seen a few articles on the subject, e.g. How pure are the colors of the rainbow?. But I haven't yet found published results of narrow-angle, calibrated photometry on rainbows in either optimal-contrast or average viewing conditions.

Can someone here please point to such an article, preferably not behind a JSTOR-type paywall?

  • $\begingroup$ Do you require measurements from real rainbows, or would a calculation from a simulated rainbow suffice? The latter is not too hard to do using the Mie theory (though do note that saturation of a rainbow's colors depends strongly on the distribution of drop sizes, see e.g. this page). $\endgroup$
    – Ruslan
    Commented Aug 12, 2022 at 16:48
  • $\begingroup$ Thanks for your reply, Ruslan. "Require" would be an overstatement - it's just long-term curiosity. I have seen some good web pages that offer Mie calculations, as you provide linked, sometimes mapped onto various color spaces. What I'd really like to see, though, is actual measurements - especially since so few people use calibrated monitors, so on-line simulations are suspect. Seems that someone must have made these measurements, if only out of curiosity. I'm really seeking actual color saturation or chrminance values, which can then easily be compared to, for example, Web colors. $\endgroup$
    – cTen
    Commented Aug 12, 2022 at 17:09


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.