I'm not sure if this is on topic, but here goes.

Cyan in RGB is (00, FF, FF). Red (or green or blue) is this (FF, 00, 00). So this seems like if I tell my monitor to display cyan, each pixel will have 2 subpixels fully activated. If I tell my monitor to display red, it will only have one. So a monitor with a peak brightness of 300 nits would only display 100 for red and 200 for cyan.

Cyan in HSL (or HSV, is there a difference?) has a hue of 0.5, saturation of 1, and a lightness (or value?) of 1. Red has a hue of 0 (or 1, these are identical, right?), saturation of 1, and a lightness of 1. These both have a lightness (or value) of 1, so the 300 nit monitor displays 300 nits for both colors.

So which (if any) is right? The second one makes sense to me, I wouldn't expect cyan to be any brighter than red, but I can't figure out why the first doesn't make sense. Am I describing color completely incorrectly?


As suggested, I'll edit to include how this affects my workflow. Short answer, it doesn't. But I think understanding color in general is very important to an art which is literally just capturing color on an xy grid. If I had a question about how the aperture actually opens or closes on a lens, I'd ask that on a photography network, not an engineering network.

  • 4
    \$\begingroup\$ What does this have to do with photography? \$\endgroup\$
    – OnBreak.
    Commented Apr 15, 2020 at 4:49
  • \$\begingroup\$ You might want to add how this affects your workflow in photo editing, otherwise this borders on being off-topic. \$\endgroup\$ Commented Apr 15, 2020 at 5:39
  • 2
    \$\begingroup\$ I’m voting to close this question because it doesn't have an obvious relationship to photography. \$\endgroup\$
    – Philip Kendall
    Commented Apr 15, 2020 at 7:55
  • \$\begingroup\$ Dark amber red or Ferrari red ? \$\endgroup\$
    – Alaska Man
    Commented Apr 15, 2020 at 17:47
  • \$\begingroup\$ @AlaskaMan (255, 0, 0)... Pure fully saturated red. That's in the question. \$\endgroup\$
    – user91535
    Commented Apr 15, 2020 at 17:53

2 Answers 2


The RGB values for cyan are 0,255,255 RGB (8bit), and red is 255,0,0. In RGB, if any of the max values are the same then the brightness is the same; and the other two values determine the hue. E.g. 255,0,0 (red) is the same brightness as 0,255,0 (green) and 0,255,255 (cyan).

HSL/HSB/HSV are also not good for direct comparison as the saturation and brightness values are also relative to the hue and not to each other.

I.e. any value of 255 or 100% means the color is as bright as it can be, but it does not mean the color is lighter/brighter than another color.

enter image description here

If you want to compare them for luminance, it's probably easiest to use the Lab color space; where red has a luminance value of 54% and cyan has a luminance value of 91%. I.e. cyan is a lighter/brighter color.

You will only get the maximum brightness from the monitor when all RGB pixels/sub pixels are at max value; i.e. displaying a white screen. When displaying a cyan screen it is less bright, but brighter than a red or black screen.

And similarly, you cannot assume that any two color pixels will generate the same luminance because they are weighted differently; perceptually, yellow/green is the most important. E.g. red (255,0,0) has an L value of 54, whereas green (0,255,0) has an L value of 88. And yellow (255,255,0) has an L value of 98 compared to an L value of 91 for cyan (0,255,255).

  • 2
    \$\begingroup\$ Why do you say I shouldn't use hexadecimal? AFAIK, FF in base 10 IS 255. At the start of your answer, you say that the max value for RGB is all that matters for luminance, but then the last paragraph says that a monitor IS brighter when displaying cyan over red. Is there a difference between brightness and luminance? \$\endgroup\$
    – user91535
    Commented Apr 15, 2020 at 16:57
  • 1
    \$\begingroup\$ I guess you can use hex if you know all of the values and how they relate; I don't... RGB, HSL, etc, only specify "brightness" in relation to the hue. Only LAB separates lightness (L) in a relative manner that is easily compared like you want. RGB can be used similarly except that the value/importance of green is disproportionate and red/blue are different as well. There is a difference between the brightness of a color and the luminance of the monitor/color. I.e. if you take white light and remove some of the electromagnetic energy (color spectrum) what remains cannot be just as bright. \$\endgroup\$ Commented Apr 15, 2020 at 19:18
  • \$\begingroup\$ Hex is literally just a different way to write the numbers. It's not really complicated — in base 10, "255" means "2 hundreds plus 5 tens plus 5 ones". In hexadecimal, the base is 16 and letters A-F represent numbers ten through fifteen. So, FF is "15 sixteens plus 15 ones" — which is 255. \$\endgroup\$
    – mattdm
    Commented Apr 16, 2020 at 6:22
  • 1
    \$\begingroup\$ I suggest you remove this part of the answer as it detracts from the rest. \$\endgroup\$
    – mattdm
    Commented Apr 16, 2020 at 6:23
  • \$\begingroup\$ @PleaseReadMyProfile, I've removed the hex reference and edited for better clarity. \$\endgroup\$ Commented Apr 16, 2020 at 12:24

See Wikipedia

The difference between HSL and HSV is that in HSL, only white can have L=1. Saturated colors are always L<1.0. In other words every triplet in HSV can be mapped to RGB, while many HSL values have no matching RGB (out of gamut colors).

In the HSV model, Cyan is (.5,1.,1.) V is max(R,G,B) but in HSL, Cyan is (.5,1.,.5), L being the mid point between max(R,G,B) and min (R,G,B).

On the other hand the perceived luminosity is a different beast, because our eyes are much more sensitive to green so the real luminosity is closer to the "Luma" used in the YCbCr or LCh color models. Not everybody agrees on the actual coefficients, but they all give more importance to green, and much less importance to blue, for instance HDTV standards use:

Y' = 0.2126✕R + 0.7152✕G + 0.0722✕B 

You can check that the sum of the coefficients is 1, so only white has L=1.0, and any other color is darker. With this formula, Red is L=0.2126 when Cyan is L=0.7874.

  • \$\begingroup\$ The numeric values in Y' = 0.2126✕R + 0.7152✕G + 0.0722✕B applies after the image is decoded to be linear. See en.wikipedia.org/wiki/Relative_luminance . But the values we see in our photo image files and histograms gamma encoded \$\endgroup\$
    – WayneF
    Commented Apr 15, 2020 at 15:18
  • \$\begingroup\$ Yes, but since the examples only use values 0 and 1, the final value is still correct :) \$\endgroup\$
    – xenoid
    Commented Apr 15, 2020 at 15:34
  • \$\begingroup\$ You said some HSL values are out of the RGB gamut. Can those colors even be displayed on a monitor, then? If they can't is there a point to using HSL in photography or other digital artforms? \$\endgroup\$
    – user91535
    Commented Apr 15, 2020 at 17:05
  • \$\begingroup\$ (.5,1,1) Would require G and B values to be brighter that 1. So that depends where your white point is wrt to your display luminisoty. The point of HSL is that is closer to how we see colors. In image processing it is much easier to predict what happens with a "Luminosity" slider than with R, G, and B sliders... \$\endgroup\$
    – xenoid
    Commented Apr 15, 2020 at 17:15
  • \$\begingroup\$ @xenoid I imagine the reason using linear is stressed as important is that while the coefficients do add to 1, the gamma RGB colors are nonlinear values that determine the luminosity. For gamma 2.2, gamma 255 is linear 255, but gamma 128 is linear 56. Not using linear produces very different numbers. \$\endgroup\$
    – WayneF
    Commented Apr 16, 2020 at 15:45

Your Answer

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