You can't believe your eyes: there's no red in this traffic light

I guess from a technical standpoint yes, but I would say the average person would pick out darker colors moving along a gradient scale when asking to make a color “darker” or “lighter”. From a paint perspective part of it adding more pigment to make the color deeper or more intense, but a bigger part of it is understanding what pigments you have to work from and how they interact. The Behr paint system has multiple pigments in various ranges.


The best paint person my wife and I ever came across had a full time job as a comic artist. We could tell him the direction we needed the color to shift and then by looking at the pigments available he was able to shift the color pretty accurately. Most paint counter techs think when you “scan” a color you’ll actually end up with a paint that is that color. I’ve found that scanned colors are pretty lousy at matching.

(Paint techs typically hate us.)



My entry for the Pendant badge:

Color is absolutely relative, especially for prints (CMYK) and monitors alike (RGB). (Those are two separate color rendering systems, by the way.)

It depends on the white balance of the light being shown on the surface (for prints, wall paints, etc.) or in the light the monitor is using for backlight for LCD panels. I can’t really speak on OLED screens, that’s outside my knowledge.

It really gets fun an entertaining when you start dealing with printers capable of cranking out proofs for pre-press, because so much of it also depends on what the print driver’s feeding the printer as far as color information. (which also depends on what the program feeding the printer driver is doing for color information; Fiddle with any of those things, and the color output will change. It’s hair-pulling excitement for the poor service tech that has to deal with the service call of ‘printer not printing proofs with correct color’ only to find out that when printed using method A it comes out this way, and with method B it comes out with enough of a subtle difference that the pendant who opened the ticket can tell… (I was, in fact, that service tech many years ago.)

1 Like

And now you aspire to be that pedant who can tell the difference! Excellent!! I’d give you a badge if I could. :wink:

1 Like

No red light passes through a perfect cyan filter. It’s not clear to me whether the presenter is talking about a physical filter over the lens of the camera (which would almost certainly be imperfect) or a digital post-processing filter (which would be imperfect because color information has already been lost in the initial image).

A perfect cyan filter should have turned the red to black, not gray. The white and gray highlights, should have been converted to shades of cyan. Blues and greens (the components of cyan) should have come through as themselves. If you continue to stare at a filtered image, your brain will soon attempt to color correct. That’s the effect the illusion is trying to illustrate.

Once you start with a photograph on color film or a digital camera, you’ve already lost most of the spectral information of the original scene. What’s recorded is a different spectral distribution that’s designed to appear–to a typical human eye–nearly the same as what that eye would have registered in the moment. Post-processing the captured spectral distribution (like applying a cyan filter) gives a different result than if the filter had been applied to the original spectral distribution and then captured.


Well anyway, about the traffic light…

It looked red to me until the first grey block went up across the top, and then it snapped instantly to grey.

So I rewatched the video and noticed that if I focused on the red light only without looking at the rest of the stoplight, it slowly faded from red to grey, but if I looked at the rest of the image the red came back.


This topic was automatically closed after 5 days. New replies are no longer allowed.