0

I'm have a color calibrated camera and projector pair. I'm trying to come up with an algorithm based on color additivity, and I'm wondering whether color additiveness holds under the RGB space. For example, will the colors in RGB with values of (30, 30, 30) and (60, 60, 60) produce colors close to (90, 90, 90). I'm observing that this isn't the case and it produces a color like (72, 72, 72). I'm wondering whether this is due to some system error, or do I have to go into a different color space like YUV or Lab? Or whether I'm misunderstanding color additiveness and the color additive property does not apply for the addition of separate color components?

EDIT: I'm talking in decimal values.

dev_nut
  • 2,476
  • 4
  • 29
  • 49

1 Answers1

0

Usually RGB integer values in domain [0-255] are 8bit values with the working RGB colourspace OECF (transfer function) applied thus non linear values. If you want to perform arithmetical operations on those values you ideally need to apply the inverse OECF to the values prior the operations.

This PDF describes the sRGB colourspace OECF for example: http://www.color.org/srgb.pdf

Kel Solaar
  • 3,660
  • 1
  • 23
  • 29