LED lighting is now commonplace across homes, businesses, and industrial settings. It uses little energy and provides a great deal of light. However, a new study suggests it may come with a trade-off. New research suggests human vision may not perform at its peak under this particular form of illumination.
The study ran with a small number of subjects (n=22) aged between 23 to 65 years. They were tested prior to the study for normal visual function and good health. Participants worked exclusively under LED lighting, with a select group then later also given supplemental incandescent light (with all its attendant extra wavelengths) in their working area—which appears to have been a typical workshop environment.
Notably, once incandescent lighting was introduced, those experimental subjects showed significant increases in visual performance using ChromaTest color contrast testing. This was noted across both tritan (blue) and protan (red) axes of the test, which involves picking out characters against a noisy background. Interestingly, the positive effect of the incandescent lighting did not immediately diminish when those individuals returned to using purely LED lighting once again. At tests 4 and 6 weeks after the incandescent lighting was removed, the individuals continued to score higher on the color contrast tests. Similar long-lasting effects have been noted in other studies involving supplementing LED lights with infrared wavelengths, however the boost has only lasted for around 5 days.
The exact mechanism at play here is unknown. The study authors speculate as to a range of complex physical and biological mechanisms that could be at play, but more research will be needed to tease out exactly what’s going on. In any case, it suggests there may be a very real positive effect on vision from the wider range of wavelengths provided by good old incandescent bulbs. As an aside, if you’ve figured out how to get 40/40 vision with a few cheap WS2812Bs, don’t hesitate to notify the tip line.
Thanks to [Keith Olson] for the tip!
Wow, who knew, adding a task light to your work place improves performance. I scrolled through the research paper and i couldn’t find the control group where they gave the people LED task lights instead of incandescent task lights. That is a huge oversight.
Also only 22 participants.
Furthermore, they appear to have used very shitty LED lighting. That huge blue pump diode peak is not there in good quality lamps. Some of it is there but the YAG phosphor output can follow the incandescent curve in the blue direction quite nicely. Some deep red may be missing, but LED chips with multiple dies (blue+YAG+monochrome red) are in development. I guess you could add IR leds to that too.
That would be the proper test if IR light improves wellbeing and/or performance. Led overhead with led task light, and then the task light with or without additional invisible IR radiation.
Right, but are “quality lamps” what most people are going to be exposed to, day to day?
Talking about “day”, where is the line for “actual daylight”?
It’s hard to find high CRI lamps in the stores because people just don’t know to demand them. They don’t know what it means, they’re less efficient and slightly dimmer, and they cost significantly more, so they don’t sell well.
When you go to the hardware store to buy a lamp, it’s all the same cheap junk, at any price you may wish to pay. You have to go out of your way to find a LED lamp with a CRI of 95 or better.
Not to mention the products that people buy out of the bargain bin… you’re lucky if it doesn’t strobe at line frequency like a dying fluorescent tube in a freezer. “Warm” colored bulbs might even cast pink because of the blue diode in combination with a deep red, missing almost everything in between. It just looks wrong, which is why many swear by the cooler “daylight” bulbs instead, which are far too blue for normal living spaces and lighting intensities.
People use terrible lights, and because the color rendering is poor, the color temperature is mismatched for the amount of light, and the light is unstable at worst, it feels like you can’t quite see properly. Like you can’t focus your eyes. To compensate, people install more powerful lamps and light their living rooms like office spaces and factories.
I would say it’s changing. I started noticing CRI ratings appearing on even cheap lamps recently. Of course the ratings are fabricated as these are cheap products from China but their presence indicates growing customer demand.
Yes, they tend to print them if it’s above 80 like everything else, and leave it out if it’s below.
In comparison, the old spiral CFL bulbs never went above CRI 85 to 87 and people hated them, for various reasons, until they banned the regular light bulb and that’s all you got. Most LED products on the market are just as bad or much worse – the only practical difference is that they don’t take a minute to warm up.
Also remember that the CRI number was printed on compact fluorescent bulbs 20 years ago as well, and the consumers didn’t get the point then either.
Some models were made with CRI above 90, but they were more expensive and nobody stocked them because they didn’t sell well. You couldn’t get a good CFL tube to compare, so the consumers never thought to ask for them, so nobody made them. Progress halted at the lowest common denominator and the lamps were cost-optimized barely above the point of the average consumer developing a migraine.
Same thing with LEDs. Phillips did the L-Prize 60 Watt equivalent LED bulb that got CRI 99 simply to win a competition and collect some sort of a subsidy, then priced it at $50 a pop, and soon pulled it off the market. They never wanted to sell them, because they wanted to cost-optimize the LED bulb for lower quality and greater profit.
This has enough flaws in it just from what Lewin posted that I didn’t even go to the article. Its entire premise seems to be that mild optical illusions (misprocessing by the brain) exist, and they can be learned to be processed out. Thats not even getting to the small sample size or light quality. This feels like it was a PhD paper, or a publish or die quota filler. I hope HaD doesn’t have a publish or die quota that caused a garbage paper to be seen as exciting…
It does read like a textbook example of the Hawthorne effect.
Whether you add lights or dim them down, the fact that you’re changing things and measuring the outcome makes people change their behavior for the better.
Ignoring the small n in this, the comparison between LED and incandescent in a workplace environment pointedly ignores heaps of data that already exists. The difference is that the LED is, in fact, just another fluorescent light (UV emitter + phosphors) and that work has been done for decades. Remember the cold, blueish glow of cubicle farms, factory floors, and waiting rooms? They went somewhat by the wayside with better phosphor mixes and that’s been continued as LEDs grew from tiny indicators to whole-building light sources. It seems that the authors never considered this.
This leaves the question of why the work was done in the first place. Usually things like this pop up at the behest of industry/trade groups or advocacy of a particular technology or product (light pipes anyone?) but Barrett seems to be a lighting uber-geek so it may be he was up for tenure. Nature publications never hurt with that.
We’ll see if he pops up selling “health lights”, “Light guided meditation”, “lighting systems for improved productivity”, or continues doing interesting and fairly diverse things. If he goes over to the Dark Side (administration) we can be assured the office will be well-lit.
Lift the ban on true blackbody radiators. Let the market decide which lamp they want to buy.
I’ll take an incandescent bulb lighting my book over this huge waste of power called “AI” any day.
That’s so true. I didn’t give up incandescent to save electricity just to pay more so someone else can waste it.
Hint: incandescent appliance light bulbs are still available. You can even find frosted bulbs, and they’ll last longer than the standard halogens, though ironically they’re even less efficient. They’re the same as “rough service” lamps – old fashioned tungsten bulbs with sightly thicker filaments and more supports.
They’re officially supposed to go in your oven or fridge, but there’s nothing to stop you screwing one into your desk lamp or ceiling light fixture. Just make sure it can handle the heat – modern ones are made with a plastic that will melt.
Some more food for thought – as our eyes age and loose our blue sensitivity, we loose much more with LED sources that simulate white light:
https://www.mikewoodconsulting.com/articles/Protocol%20Winter%202020%20-%20Our%20Ageing%20Eyes.pdf
“lose”
I wonder if this is similar to why any non warm-white & shitty “color accuracy” LED bulbs meant for replacing incandescent and fluorescent bulbs always put me in a shitty mood long-term.
Though I always bought high-end warm-white LED fixtures if I had to buy the type advertised as having non-replacable light source.
Partly because I didn’t want to replace a whole fixture if LED’s die or get sufficiently dull relatively early, or cobble up a replacement part for it if the OEM one is some weird shape or voltage/current requirement.
Whether cool white (4000k) or daylight (6500k, which is NOT sunlight – sunlight is closer to 3500k in western europe) is appropriate depends heavily on the illumination level. Search for ‘Kruithof diagram’.
The darker it is, the lower the color temperature needs to be (which is why dimming incandescent lamps is pleasant). In a brightly lit (say 500lx or more) electronics assembly line, the warm color of incandescent lamps is detrimental and unpleasant, and a 4000k color temperature feels much nicer.
6500k daylight at medium levels like 300lx in an average office particularly unpleasant. Everything feels gray and cold. But up that level to 800lx and suddenly you have a refreshing activating atmosphere.
This is also why 6500k LED street lighting in residential areas makes it feels so bleak. At those levels, the 2200k golden white/yellow of old school high pressure is more appropriate. Good municipalities use 2700k LED lighting in residential areas, which is really comfortable. It’s dimly lit, but feels cozy.
I use both 2500k, 2700k and 4000k >90CRI LEDs at home. The 4000k ones i use in the daytime in autumn and winter to remove the gloomy sensation of the everlasting dutch cloudy weather. 4000k matches the outside light with heavy clouds very well.
2700k and 2500k are used in the evening with low light levels in the order of 75lx.
I agree that the narrow spectrum peaks of LED lighting may cause some perception oddities.
However, given that most of us are exposed to broadband lighting at some point during the day (incandescent or sunlight), the claim that exposure thereunto shifts color sensitivity in a lasting way makes me very skeptical about this set of claims.
We do know that color perception renormalizes to try to compensate for the lighting available; the Land effect. Under orange halogen street lights, a red car may be seen as grey, as we try to subtract out the overall lighting bias. It is possible that seeing specific objects under broadband light allows us to recalibrate our expectation of, and therefore perception of, their colors.
I do have to say that replacing the old fluorescents in my workshop with “daylight” LED fixtures made the space tremendously more pleasant for me. But that may be because I also nearly doubled the lighting level.
I will also point out that historically switching offices from incandescent to fluorescent lights produced both positive and negative reactions from those working in these spaces. (As evidence, I cite the lyrics of the song “Coffee Break” from the musical How to Succeed In Business (Without Really Trying), where fluorescent lights, though desirable, rank below coffee.)
I do like decent fluorescent lighting. But one thing that LED’s don’t but fluorescent lighting does is HUMMMMMMMMMMMmmmmm. Okay, I know it not the light that does it, it’s the ballast but still, it’s pretty noticable. On the other hands, it doesn’t distract from the BBBBBBBRRRRRRRRZZZZZZZ of the coffe machine, the WHEEEEEEEEEEEEEE of the fans in the patchpanel cabinet or the RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRR coming from my PC fan. Sure my co-workers complain about my loud typing YTERWYTVWYTBFWYTBFWYTBFWYTWB but that does not seem to bother me personally and it keeps me awake preventing me from falling… ZZZZZZZZZZZZZZZZZZzzzzz ZZZZZZZZZZZZzzzzz ZZZZZZZZZZZZZZzzzz
Considering most indoor work environments at the moment are lit with either LEDs or florescent lighting, I’m not sure what the relevance of these findings are.
{trollmode} simple. solve it with MOAH LIGHT like they do with cars. that makes everything better… {/trollmode}
Bring back tallow candles
It might be hard to imagine some future government with kooky ideas about technology and health coming to power, but bear with me: a country where masked police roam the streets with spectrometers looking at people’s illumination at night and kicking down the doors of those not using “proper patriotic” bulbs!
The future is here and now.
… and arrrressssting those deemed “obtaining bulbs from unfriendly countries” …
We already had that – when the government told you you couldn’t have incandescent bulbs anymore.
They were very prescient, those technocrats – they knew we were going to need to save money for Silicon Valley’s AI datacenters so people could generate pictures of politicians in bikinis.
LED lighting is poorer for task lighting, just compare to a halogen lamp and you will notice SMT rework is so much easier.
When I researched the “why” behind this, the papers said high (long wavelength) red spectrum is essential. So warm white LED’s, no flicker as well, are much better than “pure white” malarkey.
This sure looks like the Hawthorne effect – Wikipedia https://share.google/s46k6WX737PPEPxIU
Yeah I thought the same. I didn’t read the study though, Hawthorne effect is so classic that I would have expected them to consider it.
When you have a predetermined goal you only worry about the effect if it’s negatively impact your goal.
Isn’t it 20/20 vision you’re looking for, not 40/40?
Would there exist some kind of filters toning down the blues AFTER?
Considering the short IR exposures needed (3 minutes already had an effect), maybe we should just go outside during the day.
IR could readily be added for evaluation with LEDs.
Then you might not be on your phone!
That graph of spectrum is one of the most deliberately misleading graphs I’ve seen in the past several years.
We can’t see anything outside the 400nm to 700nm band. Pretending that the huge swath of 700nm and longer is relevant can only be there because the authors are either ignorant or trolling.
780nm is visible if it’s bright enough