This is interesting commentary on an interesting article. Will there be a future mass market for displays that are better than RGB? If so, what will drive it?
Originally shared by Sam Vilain
I’ve known since a 100–level physics paper about the horseshoe diagram, but never looked into how it was derived—and how a 3–dimensional space of possible excitation levels of cone cells is mapped into a 2–dimensional graph, and why monochromatic light ends up making a curve on it. I’m still not quite sure how additive primaries end up making a neat triangle on this graph, but it’s interesting stuff. I liked the idea that to express certain colors you need “negative” RGB values; that requirement shows the areas impossible to reach using your chosen phosphors.
It also makes me wonder, with the obsession with things like “Retina Displays”, why no display manufacturers are differentiating by adding more primaries, so that their displays will be able to cover more of the full horseshoe of visible colors than their competitors. They could interpret the RGB values in a conventional display buffer into a larger triangle, and then map the output to 4+ primary phosphors; probably (at least) more in that Green-Blue area of the horseshoe diagram which is always cut off to the left of the triangle.
xvYCC tried with little fanfare. Sharp had their 4 color (RGBY) panels. It’s a hard technology to introduce because there is no backwards compatible way to broadcast it. And most consumers don’t care… I suspect high framerate and high dynamic range will catch first.
Dennis Adams, thanks. I hadn’t heard of xvYCC. I just read the wikipedia article about it, and it sounds like a very rational, mostly-compatible step forward.