[ELI5] What would happen if instead of using just red, green and blue, we used more pixels colors, like yellow or white?
In: Technology
The big problem with this is it would take more bandwidth for us to transmit images.
Using RGB color, you’d do something like #dddd88 or the bits 110111011101110110001000 to send a light yellow. White is #ffffff, or 111111111111111111111111. (3 colors, 8 bits each, 24 bits.)
If instead of that you had “red-green-blue-yellow-white” color, light yellow might be something like #000000ff44 (yellow plus some white to lighten it up) and the bits would be 0000000000000000000000001111111101000100. (5 colors, 8 bits each, 40 bits now.)
White, in this system, would be #00000000ff which would again require 40 bits.
We’d be using more bits, which means either more time or more parallel wires, to produce a color that we could make in the RGB system anyway. Video processing would either go a bit slower, or cost a bit more because of the added hardware requirements. Not sure that’s worth it.
If you really want to make monitors better – check this.
https://upload.wikimedia.org/wikipedia/commons/thumb/6/60/CIE1931xy_CIERGB.svg/325px-CIE1931xy_CIERGB.svg.png
The colors inside that triangle are the ones we can make using RGB combinations; the ones outside the triangle, we can’t hit accurately. (That is, that image isn’t showing them accurately either.) So there are some colors in the pink-violet area that an extra pixel could help us access, and there’s a whole bunch of action in the cyan/turqoise area that we’re missing out on.
Human color perception can be mapped out in a 2-D graph: https://en.wikipedia.org/wiki/File:CIE_1976_UCS.png
As an engineer, you’ll pick real-life materials that produce your 3 chosen primary colors (in real life). On the graph, you’ll plot the 3 points corresponding to your primary colors. You’ll connect those points with lines, to make a triangular “gamut.” The gamut encloses all possible colors that you can recreate from combinations of your 3 chosen primary colors.
So, there’s no need to have an extra “white” pixel, seeing as the “white” is already in your gamut.
Instead, you could consider searching new technologies to make very intense hues at the outer boundaries of the color graph, to allow a wider gamut.
Not a whole lot.
A white LED subpixel would make a brighter white than the mixed RGB, a yellow subpixel has been experimented with, and technically adds extra “colors” available for the eye, but most people can’t see them as different from colors made by RGB.
And piling in too many subpixels in a space makes the pixels too big, losing resolution compared to just the three needed for RGB color.
There would be no difference because RGB makes up all of the colors visible to the human eye.
The fact is, human eyes lose information. A wide variety of possible combination of mixture of wave length and intensity get mapped into the activation pattern of just a few type of cells. Because of this, completely different mixture of light can produce the same color.
If your goal is merely reproduce the image as how a human would see it, all you need are a few pixels that activate these cells the same way. RGB is based on our understanding of what cells are there in our eyes.
Latest Answers