| Reference | Downloads | Github

Colors with same luminosity



Hi everyone!

I have the following situation:
I want to present barplots as stimuli in a EEG experiment.
To controll for perceived brightness, all three colors used (red, blue and the gray background) should have the same luminocity.
I am somewhat convinced that my red and blue came out the same in brightness, but the gray background looks darker than the bars. I haven’t had the chance to test it in the lab, but that’s the impression I get on my home-monitor.

Using the luminocity weights by the WCAG 2.0 (Y = 0.2126R + 0.7152G + 0.0722*B) i figured out the following three colors to use (the values are in psychopys system reaching from -1 to 1):

Red (just pure red): [1, -1, -1]
Blue (With a bit of green to get the same luminocity: [-1, -.6073, 1]
And Gray (where all three values should be the same): [-.5748,-.5748,-.5748]

The screen is a Samsung SynchMaster S24B350 LED and I’m runing Psychopy 1.84.2 (standalone) on Windows7.
Do I need to apply some additional gamma correction or something similar? Or is my result correct and I can’t trust my eyes? :wink:
Does someone have some more experience with this sort of problem?

Thanks a lot!


There’s just no way of knowing, without calibration equipment (a spot photometer) what colors will be equiluminant on a given monitor. You have to measure on your specific device the relative brightness of each gun. If you don’t gamma correct then you also can’t extrapolate from one color to another (although you can still manually determine what colrs are equal of course).


Hey jon, thanks for your answer.
So you would suggest that I measure the luminance of pure red, green and blue on the screen, and then choose the colors by weighting them twice: Once with the measured relative luminance and again with the WCAG weights to get the perceived luminance?

Do modern screens also apply some sort of alpha correction which would render the displaying of the color strengths non-linear? I’m sorry if this is a dumb question, but the literature I found in my opinion is not very intuitive.

Thanks for the help.


I’ve never heard of alpha correction. Gamma-correction is what you would need to do to make your screen linear. I think you need a local mentor that knows about monitors and calibration though - there are too many potential pitfalls for me to teach you this remotely.


Sorry, I got my greek letters mixed up. Of course I meant gamma correction.
This toppic seems to be quite a rabbit hole to dive into.
I might know someone who may be able to act as a “local mentor”, but thank you for the help so far.