I don't know either, but if we visualize the RGB color space as a triangle that is entirely contained within the weird shape that represents the set of all colors the human eye can perceive ( https://en.wikipedia.org/wiki/RGB_color_model#/media/File:CI... ), presumably the idea is to cover more of that human-perceived space via a quadrilateral with four anchor points rather than a triangle with three. Presumably the "C" in "RGBC" stands for cyan, and in the linked image the cyan portion of the color space is particularly poorly represented.

Cyan is severely under-represented by monitors, so the extra pixel is a dedicated cyan. It dramatically improves the ability to display blue/green colours.

*edit: found the link I was after on this: https://moultano.wordpress.com/2026/06/19/where-to-find-the-...

We have a very "fuzzy" visual perception. I remember seeing an RGB response curve of the human vision mechanism once. I doubt it was measured. Maybe they extracted it from the CIELab stuff.

Anyway, things like the green (or blue -can't remember) receptor have a strong curve in the green spectrum, but also a "bump," over in red (I think).

We're an organic mess.

Looking at RGB curves for LEDs, they are three perfect little mountains. No "bumps," anywhere.

I guess that the goal is to try to mimic the "messy" human visual perception.

Also, expect these monitors to be non-cheap. Companies like Eizo are having a difficult time, justifying their prices, these days.

The cones are not sensitive to a single wavelength but to a range.

The green-sensitive cones overlap with the red-sensitive cones, and to a smaller extent also with the blue-sensitive.

Full saturation red and blue are possible by emitting light on the edges of the visible spectrum.

Full saturation green, however, also activates the red and blue cones.

To cover the whole gamut is impossible, but you can approximate it with ~three green tones: a 490nm deep cyan that hits blue and green but not red, ~510nm that hits red and blue equally, and ~540nm the peak of the green cone.

I don't have an answer, I'm just wondering out loud.

Cone cell activation is complicated. Displays with three well chosen primaries are economical and effective, but they aren't intended to produce every perceivable color. And our chromaticity diagrams, that pointy splotch that's often used to compare display gamuts, is based on a "standard observer" that is a simplified model for human perception.

An ideal pixel would be able to emit any kind of electromagnetic radiation of any intensity, kind of fun to think about but unrealistic and impractical.

What additional primaries mathematically do is expand a gamut from a triangle to a convex polygon. While ten or a hundred primaries would be bonkers, I bet we could fit a quadrilateral or a pentagon to the perceivable gamut in ways that'd see some gains.

Remember that light is synthesized by combining the primaries; the spectrum is defined by their convex hull. You can expand the hull and ergo gamut by adding primaries.

Remember that light is synthesized by combining the primaries; the spectrum is defined by their convex hull. You can expand the hull and ergo gamut by adding primaries.

The RGB setup we have strikes a balance between cost and visual quality. If the cost of adding primaries goes down you can add more to increase the quality. One issue is that the signals often assume RGB (channels), so the hardware manufacturer would have to adapt the RGB signal to their multi-primary hardware.

Humans can see more than the colors they can make with only combining RGB pixels; you can't make 'neon' colors with them, even though we can see them in real life, for example. Other commenters pointed to links showing the visible color gamut vs the RGB ones. Compare also with CMYK used in print, it can produce sightly different colors compared to display RGB.

You are probably familiar with the horseshoe-shaped chromaticity diagram [0] of human-visible colors. A light source with three color primaries spans a triangle in that coordinate system. To cover the whole horseshoe, the three vertices would need to be considerably outside the horseshoe. With four color primaries, you get a quadrilateral that makes it easier to cover a larger portion of the horseshoe.

[0] https://upload.wikimedia.org/wikipedia/commons/1/1e/CIE1931x...

It's not as simple as "3 cones = 3 primary colors." Each type of cone has a response curve and three curves overlap: http://hyperphysics.phy-astr.gsu.edu/hbase/vision/colcon.htm.... And each cone has different sensitivities (blue is much more sensitive than red and green). So perfectly monochromatic light will stimulate two and usually three cones to varying degrees. When you mix "green" and "red" to get yellow, what you're actually doing is stimulating the green cones (but also the red cones) and the red cones (but also the green cones) in relative proportions that your brain perceives as yellow. But it won't necessarily give you the exact same response of the two cones as monochromatic yellow light.

The "sometimes 4" is quite fascinating: https://en.wikipedia.org/wiki/Tetrachromacy#Tetrachromacy_in...

Yeah. That was crazy. I've not encountered that, before.

Turning off Javascript helps.

If you’d read the article, you’d learned that these are advancements in backlit LCD technology. More generally, however, having more than three color primaries is orthogonal to the question of backlit vs. self-emissive.

I remember George Takei shilling those things.

Probably yes, but density hasn’t been an issue for over a decade. Mobile device screens are 5-10x denser than TVs, and cost <€100; Apple has used mini-led for their retina displays for a long time too.