While analogous colors are the same in both the RYB substractive model and the RGB additive model, complementary colors are different.
For example, the complementary color of red in RYB is green whereas it’s cyan in the RGB model.
In RYB, blue and orange are complementary. In RGB, blue is complementary with yellow.
Also, I note that the RYB has more color space allocated to the red-yellow spectrum (5/12) whereas the RGB has less (3/12).
On the other hand, the blue to green spectrum occupies about 3/12 of the space in RYB but 5/12 of the space in RGB. Which most accurately follows human perception?
As far as I can tell, RGB is most analogous to the way light is perceived by us. For example, stare at some red object for a bit, then look at a white background. The afterimage is cyan, not green.
RYB came about before we were aware that light’s primary colours were RGB.
All of what I said here has been regurgitated from this site.
Color hue is encoded by retina into two signals: Red-vs-Green and Blue-vs-Yellow. IOW, there are four corner colors, just what you’d expect since chromaticity is two-dimensional.
The bottom line is the traditional paradigm of 3 primary + 3 secondary colors, with one additive system and one subtractive system are mostly BS physiologically and psychologically.
Like a lot of other 19th century physics and biology they’re logical and make sense; they’re just nothing like what actually happens in the real world or the human body. The real world is vastly more subtle than what “stands to reason” in 1855.
Yes, but from an artistic POV the traditional color wheel works well for the purpose of Art, mixing and matching colors (aesthetically), blending and combining paint/pigments, and in traditional (non digital) graphic design.
In human perception, neither model represents how our visual system works, but RGB/additive color is closer.
Also light vs. dark. Each level of the visual system operates a bit differnetly, all using input from the RGB-esque cones.
And these colors aren’t what you think of as RGBY. The “ideal” blue is more violet. The yellow is a gross puke green. The cones also don’t respond best to what most of us would consider a RGB.
Does it even make sense to talk about complementary colors? I’m trying to see how wavelengths or frequencies could be complementary and I’m not seeing it. Yet, colors like pale blue and bright orange do seem to have a complementary effect.
Complementarity (if that’s a word) has to be an artefact of how we see colours - simplest way to think of it is that those parts of the retina that are stimulated by a colour are exactly those parts that aren’t stimulated by its complement.
We have had threads and threads on this and closely related topics. And we have several posters who are professionals in this area who’ve made some excellent posts. Searching out these threads will pay dividends to the curious.
Not really. We do have three cones, not just two While it is true that there’s a large overlap in two of them, the S-type or blue cones really don’t extend to yellow. You can see the response curves on this chart.
That doesn’t mean there isn’t an argument for a four colored colorspace, however. This is a fairly good illustration of how colorspaces map to human vision. You can see making a quadrilateral that could capture more colors than any triangle, as long as both can only produce actual colors and not ones we can’t see.
RGBY is only an idea because yellow allows more specificity in the green, and we can distinguish green better than other colors, probably because we have two types of cones that cover that range, and our rods even respond better in that range. That’s why digital cameras use RGBG tiling.
I’m not sure what you’re objecting to. Yes, three cones provide 3 freedoms (rods, the fourth sensor type, are saturated and almost useless in high illumination); the three freedoms (i.e., dimensions) comprise one luminance dimension and two chrominance dimensions. Denoting the cone outputs as R,G,B, the retina outputs are, approximately, weighted sums of R,G,B:
Luminance = R + G
C[sub]r[/sub] = R - G
C[sub]b[/sub] = 2B - (R + G)
I’m not sure either, as the “good argument” for what septimus is talking about is “tons and tons of actual physiological research.” Yes, the cones have 3 primaries, but you don’t even leave the retina and that signal is transduced to a system of 3 opponent signals.
