Yellow to Blue: the Recent History of Lighting and Color

A recent LA Times article that calls attention to the increasing prevalence of blinding headlights on the road at night got me thinking about why this is happening and what it means for the near future of lighting. Maybe it’s just my aging eyes, but at times I find myself flashing oncoming cars who appear to have their brights on but don’t, because they’re too bright and blinding. And the encroaching illumination of roadways with high blue content glaring LED lamps is only exacerbating the situation. These two developments are connected, because as lighting is getting more energy efficient and the cost of new lighting technology is dropping, light is getting bluer. This is often generally not a good thing, for many reasons. But the most important is that (IMHO) we may deeply hardwired for warmer light at night, and the juggernaut of blue light technology seems to be happening in total ignorance of and opposition to our basic physiology.

Unfortunately, with LEDs, the prevalence of high-blue light is, like many things in the history of technology, something of an accident. Here’s an abbreviated bit of the history behind this. The red LED was invented by Nick Holonyack in 1962, and yellow and green LEDs soon thereafter. But a viable blue LED remained elusive for several decades. The tipping point for the commercialization of LEDs came when Shuji Nakamura, Isamu Akasaki, and Hiroshi Amano developed a blue LED based on gallium nitride deposited on a sapphire substrate. Red, green and blue LEDs were necessary to create white light. Commercial LEDs today make light by producing what’s called a “fundamental emission” which in most devices is a pure blue light that registers around 450nm on the electromagnetic spectrum. (You knew that, right?) This pure blue light then excites phosphors embedded in a “package”- a silicone disk deposited directly on the diodes, or located close to them. These phosphors contribute red and green light, which when combined with the fundamental blue, produce white light. Thus, because of the underlying chemistry and quantum physics behind the materials Nakamura and others were using to make the early LEDs, most of them today produce a spectrum of light that is high in 450 nm blue. Other LEDs were developed, notably by Soraa (founded by Nakamura), that use a violet fundamental emitter, and a blue phosphor to create the requisite blue in addition to red and yellow or green, but this is not the dominant LED lighting technology.

So today, as electric light technology marches on, getting more efficient and bluer all the time, we are all looking back nostalgically at our old-fashioned, wildly inefficient warm incandescent light sources, (although we’re not looking back with warm feeling at fluorescents, as we never really liked them in the first place). This looking over the shoulder is quite consistent with the short history of “artificial” lighting. Gas lamps were decried as too bright and glaring (as light gets brighter we seem to see it as bluer) compared to firelight, and as they became more efficient they indeed became brighter and bluer. Then incandescent electric lights were seen as too bright and too blue compared to gas lamps. Then fluorescents had the same reception, although because of their technical shortcomings they didn’t even come close to completely replacing incandescents, partly because of their deficient color rendering and poor performance (which true to form, improved rapidly just as they began to decline in popularity, which is exactly what gas lamps did).

What’s different today with LEDs is that they offer the opportunity to provide a much better quality of light spectrum that, at its best, is infinitely more controllable and aesthetically pleasing than incandescents, it’s just that most people don’t know this yet because the early products on the market have been so poor in quality and the high quality products are still seen as specialty items. Another key difference is that we see that we won’t necessarily have incandescents to fall back on, as we did with the introduction of fluorescents because incandescents are being quickly phased out everywhere and replaced with LEDs.

One could easily see this as the inexorable march of technological progress- that blue is associated with modernity, efficiency, and glass and steel utopias where one can be smoothly whisked from one hedonistic episode to the next in a personal self-driving spaceship. And our fond romantic association with candlelight, tiki torches, Edison bulbs, and now even HPS streetlamps, may soon be viewed as mere atavistic pining. But I think it’s not, and you probably don’t either. The picture is infinitely more complex of course, it always is.

It turns out that there is historical precedent for the fashionability of color, blue specifically, in a completely different part of history- painting and textiles. My excellent mentor, friend and colleague Gere Kavanaugh alerted me to a current exhibition at the Norton Simon Museum in Pasadena (where in the 60s as a boy of 12 I was exposed to the work of the Bauhaus, an event that had a huge impact on me): A Revolution of the Palette: The First Synthetic Blues and their Impact on French Artists. Gere knows much more than a thing or two about color, having been a color designer her whole career. She has influenced how I see color tremendously, and shares with me a passion for the history of technology. Over dinner recently in LA, she told me a bit about the story behind the exhibition and the history of how blue pigments in late 18th and early 19th century Europe transformed art, textiles, and many other things in surprising ways.

I began to think about the interplay of pigments and lighting from the standpoint of the history of technology. At the turn of the 19th century, lighting was just beginning to change from candles to gas light, but hadn’t quite yet. Artists, painters specifically, were really the equivalent of web designers today, in the narrow sense that before photography they were in charge of graphic communication in color. And they did not have a full compliment of the colors we take for granted today, they were specifically missing good blue, just like LEDs in the end of the 2oth century. Of course association of specific colors like purple with high status, has a long history, as in ancient Greece and Rome. Our perception of color is malleable and greatly impacts how we cognitively negotiate our environment. The commercial development of three specific blue pigments- Prussian, Ultramarine, and Cobalt – coincided with the Enlightenment, and greatly affected the evolution of the color theories we use today to understand light and color.

View fullsize

As public lighting blossomed in American cities in the decades before and after the turn of the 20th century, artists like Georgia O'Keefe (left) and Alfred Stieglitz (right) began to look at its transformative effects in different ways. Nighttime photography was technically difficult, a fact that limits the historical record of early electrification. Painting was also more difficult but could convey effects of lighting that photography could not. In these images we see the effects of the Great White Way in New York, streets generously bathed in bright, glaring white (which is to say relatively bluer than gaslight) electric light. 

Could we be in the middle of another blue fad, this time with lighting? It’s possible. If this was true, all the recent talk about circadian effects can be seen from a different angle. Most of us in the lighting world understand to some degree how blue light impacts melatonin production and sleep, and we all cite the need for more research, but I often hear the sentiment expressed that it’s fine to hate overly blue light just because it’s ugly (I agree). I very much want to reevaluate beauty as a valid criteria for design decisions, especially when they effect people as much as lighting does. As a very general rule, we view as beautiful that which affords health, safety, utility, food, or improved chances of passing our genes along to the next generation. Beauty’s not an abstract concept, it’s cognitive shorthand for what to seek and what to avoid- a vital evolutionary adaptation.

What’s fascinating about the evolution of blue pigments is that it afforded artists and textile manufacturers a full spectrum approach to increased complexity in reflected light in paintings, fabric, and other artifacts. Their increased complexity was an additive process- developing more and more fundamental pigments based on chemicals extracted, processed and synthesized. And all of their perceptions and experience took place under daylight or light on the blackbody curve. Today we’re dealing with the question of spectrum in generated light, which of course affects all reflected light, indoors at least. After incandescents, our recent technology development in lighting seems to be going in a subtractive direction, removing supposedly unnecessary parts of the spectrum depending on what chemicals and materials can be affordably produced. LEDs are efficient precisely because they drop out a very wide portion of the spectrum on the infrared part that is simply wasted heat. Several recent lighting technologies like fluorescent and metal halide deliver only selected parts of the spectrum on the theory that our eyes “fill in the gaps.” Indeed, we do adapt quickly to such spectrally inferior lighting, partly because our eyes are so resilient. And of course, like LEDs, they’re cheaper to make, so we can produce lots more crappy light everywhere, which we have did with earlier technologies and continue to do with LED.

I think though that there’s a limit to blue in light. Without a lot more research it’s safe to say that most people want warm light indoors, and light with higher color rendering. The fact is that with LEDs we don’t have to be limited to blue, low color rendering light just because it’s cheaper to produce. LED technology is infinitely more flexible in its ability to make light of any color spectrum – full, narrow band, or anywhere in between. There are of course real limitations in what can be produced affordably, but we are totally capable of producing very high quality light with LEDS – even better than incandescent – affordably at large scale. Simply because the market’s been flooded with overly blue light up until now doesn’t mean we have to accept it. Unless, of course the fad of associating bluer light with wealth, status, “modernity,” and “progress” trumps an innate desire for beauty, or perhaps transforms it entirely. Stranger things have happened.