NAB 2022: Moving beyond three primary colours

6P Color’s new Full Colour Range display is a proof of concept that aims to genuinely show more of the colours that humans can see.

This year, NAB tucked its Futures Zone away in the far reaches of the vast new West Hall, but it's still a great place to find interesting new ideas. One of those ideas looked, from a distance, like an LED video wall tile - albeit with a startlingly good ability to render exactly the sort of turquoise and cyan hues that traditional electronic colour systems find tricky.

This being the futures zone, the display we're discussing, exhibited by 6P Color, is very much a proof of concept, but it certainly demonstrates a solution to a problem that's not often discussed.

There's a beguiling symmetry between the way human eyes work, and the way colour imaging systems have always worked. It’s instinctively easy to accept that human eyes have three kinds of colour-sensitive cells (which they do) and pictures have three primary colours. The problem is, as anyone who’s looked at a CIE chromaticity diagram knows, that the human visual system is not made of mathematics; it’s made of meat which turns that nice simple three-primary system into an irregular D-shaped blob representing the colours we can see.

Or, to put it another way, most of our colour pictures have a colour gamut that’s triangular and our eyes emphatically do not. No matter which three colours we choose, there will always be some colours humans can see that a display can’t create.

Six of the best

It’s not a new problem, but 6P Color has pulled together some capital to fund new work at Baylor University into ways of representing images with more than three primary colours. Under researchers Corey Carbonara, Michael Korpi and Gary Mandle of the University’s film and digital media department, the project has looked at systems with six primaries, adding cyan, magenta and yellow (which we might properly call secondaries) to the traditional RGB.

The company refers to its system as FCR, which stands for Full Colour Range. This is, to be very pedantic, still something of a misnomer, since while the gamut of the human eye is not a triangle it’s also not an irregular hexagon. The shape of that hexagon is also limited to some extent by the available types of LED and may not be entirely ideal.

Still, in a practical sense, the system potentially has more coverage than any three-primary system could achieve. Yes, there are three-primary systems that can describe any colour in an abstract sense, as the ACES colourspace does, although it uses notional, virtual colours that can’t exist in reality or be used to build a monitor. 6P’s FCR system is designed to be buildable.

The panel on display at NAB is a test article made with RGB plus cyan emitters, and the footage is designed to demonstrate the improved ability to render hues in the cyan range. Successful as this is, it might raise fears that FCR displays would require all-new material to be shot, although that doesn’t seem to be 6P’s intent. A six-primary display can show material mastered for Rec. 709 displays or sRGB displays or Rec. 2020 displays using conventional colour management techniques just as a TV designed for modern, wide gamut colour can fall back to 709 if it needs to.

Finally, then, we might end up with something that can properly display scanned Kodachrome, or three-strip Technicolor, without blunting the edge of those historic hues. Perhaps more important, though, might be the ever-so-topical subject of virtual production. Probably, the company built its example display in the style of a 2.5mm-pitch LED wall panel because that was a more practical way to build a demonstration item than the massive complexity of constructing a custom TFT-LCD panel. Under current circumstances, though, it’s hard to overlook the resemblance to a video wall component that might be used as a virtual backdrop. 

Virtual applications

Not coincidentally, virtual production is somewhere that a multi-primary video wall panel might really score. A large part of the benefit of virtual production arises from the fact that it casts interactive light on the subject, and at the moment the colour quality of that light tends to be very, very bad. Disco-light bad. Pocket flashlight bad. Adding more primaries might help. How good it could ever get is complicated, since video displays involve some very specific engineering compromises, but it would almost invariably end up being better than a conventional video wall.

If there’s a concern, it’s that it’s hard to look at what 6P Color is doing without thinking of more or less every low-cost single-chip DLP projector, which often have additional colours (and white) in their rotating colour wheels. Other multi-primary displays have been tried, and there are various attempts to expand the range of colours used on camera sensors, with emerald or yellow-filtered photosites occasionally used. Those technologies, though, are more about sheer brightness than colour gamut; that yellow filter generally lets through a lot more light than the red or green, making for brighter projectors or more sensitive cameras.

In theory it might affect the colour gamut of which that projector or camera was capable, but that’s not really the purpose of those particular developments; it is the purpose of the 6P FCR project.

Whether 6P’s work leads to big changes in how colour displays are built remains to be seen. Although the intent of the technology is not the same, it might be reasonable to fear it might go the same way as those multi-primary sensors, which were widely viewed as not making a big enough difference to justify the extra manufacturing complexity. Conversely, a six-primary system certainly can make a visually noticeable difference. Consider also the fact that virtual production is current enough to liberate some funding on the basis that more primaries really might make film stars look actively better.

So, while the 6P idea doesn’t necessarily have anything specifically to do with virtual production, serendipitous timing might lead to something interesting.