HDR: Part 41 - Creative Technology - LED Tech Update

A technology might seem to be near maturity when design approaches that were once theoretical, and impractical to build, become easier to make and start to hit the market. LED lighting isn’t necessarily easy to build in 2021, which is why it’s still not as affordable as we’d all like, but some advanced techniques are becoming more everyday.

White-light Basics
Let's recap how we got to where we are. There is no such thing as a white-light emitting LED; the physics involved ensures they mainly emit very saturated color. To anyone used to modern color imaging, the most straightforward way to make white light would be to combine red, green and blue, but that doesn't work very well outside the realm of disco lights. Take a very saturated cyan-painted object, say, and that object might look dull and unsaturated, simply because while the light might look white to our imperfect eyes, it's actually just a combination of red, green and blue. There's no cyan in it.

So, the first LED lights used the same approach we still use now: a blue LED illuminates a yellow-emitting phosphor, and the combined output approximates white light, which works reasonably well. In modern practice, the issue is that it's hard to make any two lights match, more than that there's any serious problem with actual colorimetry.

The most common issue remains a lack of deep red, which is at the opposite end of the spectrum to blue, requiring the largest conversion from the phosphor. This is intrinsically a lossy process because shorter-wavelength - that is, bluer - light has higher energy than longer-wavelength - that is, redder - light. That's why we wear sunscreen to protect ourselves from ultraviolet (very blue) light, but other light holds no danger. It's also why daylight LEDs are more efficient than tungsten ones, because the tungsten option (inefficiently) converts more blue light to warmer light.

The color quality problem with this approach was recognized because human skin, regardless of the variety of human, looks the color it does because of the red blood beneath. Fail to illuminate it properly and the talent looks unwell, although warm colors weren't the only issue. Blue LEDs only emit a very saturated royal-blue light, and yellow-emitting phosphors don't emit anything much bluer than a sort of mint green. As a result, there's a gap in the spectrum in the place where we'd expect to find cyan and turquoise light. Trying to illuminate teal or blue-green objects with poor-quality white LEDs might mean they just look blue, a situation which might not excite the production designer.

Moving To Color Mixing
This realization made color quality a key performance metric for LED lighting, and led to a push for better technology. That's often meant better phosphors, but the most visible changes in the design of LED lights haven't been improvements in color quality; they've been about flexibility. Early changes involved combining a cooler white and warmer white, so crossfading between them would approximate a shift in color temperature. That doesn't work perfectly, since the color temperature curve on a color chart is a curve, not a straight line between two points, so intermediate color temperatures look fractionally more magenta than they should, but it's reasonably functional.

As the demand grew for full color mixing, entirely new approaches became necessary. The first color-mixing lights used conventional red, green and blue emitters, adding at least one white emitter to produce a quality white, and sometimes two white emitters at different color temperatures for better emulation of changing color temperature. One benefit is that the RGB emitters make it possible to add green to intermediate color temperatures, for a more correct result.

Some lights have tried to simulate color temperature changes using a tungsten-emitting white emitter, with blue added to increase color temperature. This works, but compromises color quality as the blue emitter is very saturated. Rather than tilting the spectrum from warm to cool, it simply adds pure blue. The problem generalizes to any color mixing light which is producing something other than white; the more saturated the color becomes, the less ideal the color quality becomes. Generally, this isn't very objectionable, because the more brightly colored light becomes - as in a submarine emergency scene with red lighting – the less we expect it to illuminate things sympathetically.

Phosphor Conversion
Some manufacturers have addressed color quality in full color mixing lights by using red or blue emitters (though most commonly just red) which are based on a blue LED illuminating a phosphor - a phosphor designed to convert all of the blue to red, as opposed to converting some of the blue to yellow, as in a white-emitting LED. Again, this conversion costs us efficiency, but it also makes for a broader, less saturated red and blue, minimizing gaps in the resulting spectrum. Yes, that means the light is visibly less saturated, with reds more a rose than a crimson, although they're often still more saturated than many TV standards can describe.

In late 2021, there is no phosphor to create primary green emitters, and until very recently, phosphor-converted blue has been difficult. Phosphors (generally) can't convert up in wavelength, from redder light to bluer light. They can only convert down in wavelength, from bluer to redder. As such, creating a phosphor-converted blue requires a violet LED, verging on the ultraviolet, which has been difficult and expensive. Manufacturer Fiilex, as part of LED manufacturer DiCon, has been able to place itself among the first manufacturers to use phosphor-converted blue emitters, creating a measurable improvement in color quality thanks to their less saturated output.

The State Of The Art
The most recent designs are increasingly starting to include other colors, or even leave out white emitters altogether. Rosco DMG refers to its own Mix technology, which includes phosphor-converted red, amber and lime alongside conventional green and blue emitters, plus a 4000K neutral white. The amber works to improve coverage of the red end of the spectrum, while the lime improves green-blue colors. Perhaps most crucially, the idea of colored LEDs with broader spectrum might prompt us to think again about creating white using red, green and blue emitters, maybe allowing us to avoid the inclusion of a white emitter altogether.

Cleverness And Compromises
Some very recent designs, such as those by Prolycht, have done that, using no white emitter. They add colors spaced in between the basic red, green and blue emitters - generally amber, cyan and lime. Combined, these colors can produce a high-quality white light, but separately they can produce a wider selection of saturated colors that are not simulated using combinations of red, green and blue. They can even create clever selective-color effects by creating a light spectrum that might never exist in nature.

Still, creating arbitrary spectrums is something of a niche interest, and most lights, most of the time, are required to produce white, and at the maximum possible power. A design based on red, green and blue emitters, plus white, might mostly use only its white emitters to do that (give or take some tiny amount of the RGB for fine adjustments). In that situation, fully three quarters of the emitters remain out of use. A light like Prolycht's, having no white emitter, might be using all of its emitters, or at least as many as it can safely keep cool. Conversely, when asked to produce a saturated red, Prolycht's design might use only one-sixth of its emitters, whereas a more conventional design might be using one-quarter, and generate more light. These are all reasonable engineering compromises, but they are compromises.

That's not the only reason that describing the power level of an LED light has always been complicated. A 100-watt daylight-emitting design will seem brighter than an otherwise equivalent 100-watt tungsten-emitting design, due, as we've seen, to the poorer efficiency of generating redder light. There are differences in the apparent brightness of light at different colors in any case, due to the differences in the sensitivity of the eye (and the cameras built to imitate it). Green light at a given power level seems brighter than red or blue, for instance. Without wanting to complicate this further, it's clear that different approaches to building color mixing lights might create devices that offer us very different things even if their nameplate ratings are the same.

The Natural Conclusion
Prolycht's light has six color channels. In theory, it would be able to create even more intricate results if it had more. LEDs are available in dozens of colors all across the spectrum of visible light and beyond, though there are naturally issues of engineering difficulty, cost, and the issue of just how useful it would be to pursue designs with higher and higher channel counts. Phosphors and LED emitters are under constant development, particularly with regard to greens and deeper blues, and things may continue to improve in the details. It's hard to imagine any vast changes in the fundamental approaches to full color mixing LED lights - though that sort of prediction has proven dangerous in the past.

You might also like...

HDR: Part 2 - Brightness Encoding

Dealing with brightness in camera systems sounds simple. Increase the light going into the lens; increase the signal level coming out of the camera, and in turn increase the amount of light coming out of the display. In reality, it’s…

Camera Lenses: Through The Looking Glass, Part I

That lump of glass on the front of the camera is often taken for granted, but John Watkinson argues that choosing and using lenses wisely can make a big difference in captured image quality.

Creating Virtual Sets With LED Walls And Unreal Engine With Andy Jarosz

Virtual production based around LED walls involves a disparate collection of technologies, and the people best placed to get the best out of the technology are often multi-disciplinarians with experience across several fields.

HDR: Part 1 - The State of HDR

Over the century or so we’ve been making moving images, a lot of improvements have been dreamed up. Some of them, like stereo 3D and high frame rate, have repeatedly suffered a lukewarm reception. Other things, like HD, and e…

2022 NAB Show Review, Part 2

With fewer exhibits and smaller crowds, the 2022 NAB Show aisles were easier to navigate and exhibitors had more time to speak with visitors.