Looking at the quality of LED light

A while ago I read what Paul has to say about incandescent light vs LED light. You can find his thoughts on the topic here.

While there is lots of good stuff in his argument and I agree with him, I struggled quite a bit with his (admittedly subjective) analysis of the quality of LED lighting. Sure I figured CFLs were terrible because I feel bad when I'm around them. That said, strangely I had never really considered that there could be an issue with LED lighting. So I spent three hours on the internet. Wow there is a lot of stuff out there on this.

There's people out there who have done the analysis and found LEDs (not the crappy ones) that have a spectrum that pretty much exactly mimics the spectrum of natural daylight. While natural light is very good for us, it is important to remember that we have evolved for thousands of years without electric lighting. Good LEDs may be just fine for us during the day, but don’t we mostly need lighting in the evening? Incandescent light produces a very warm light, mostly lower on the energy spectrum with a lot of orange and red in the mix. This is a good fit for eyes that have adapted for thousands of years to sitting next to a fire or lighting with candles in the evening. LEDs produce a lot of blue light, high on the energy spectrum.

There's a lot of academic research that has shown that blue light in the evening suppresses the secretion of melatonin, wrecking natural sleep cycles. This is the same reason we’re not supposed to look at screens right before bed. Not only does blue light affect sleep, it also affects emotional behavior, hormone production, and a whack of other things. There are dozens, if not hundreds, of published papers out there if you want to spend the time reading them all. Here's a few links:

https://www.ncbi.nlm.nih.gov/pubmed/20030543

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4734149/

https://www.health.harvard.edu/staying-healthy/blue-light-has-a-dark-side

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5299389/

Some will naturally wonder why they don’t just make LEDs that mimic the incandescent spectrum instead. After all, there are LEDs in all sorts of colors out there. Surely those smart engineers could figure it out. I am sure that they could. In fact, someone has probably already done it. Unfortunately the particular elements used in the production of blue light LEDs (from which “white light” LEDs evolved) happen to be the most economical and efficient ones around. Switching the spectrum to be more in line with that of the incandescent bulb would impact both the cost and the efficiency, which might put a damper on sales. If they could somehow figure out a way around that then there might be a hope for better quality LEDs. Until such time I will continue to stick with the incandescent as my choice for the light bulb with the highest quality of light.

Shawn Klassen-Koop wrote:...
...
Some will naturally wonder why they don’t just make LEDs that mimic the incandescent spectrum instead. After all, there are LEDs in all sorts of colors out there. Surely those smart engineers could figure it out. I am sure that they could. In fact, someone has probably already done it. Unfortunately the particular elements used in the production of blue light LEDs (from which “white light” LEDs evolved) happen to be the most economical and efficient ones around. Switching the spectrum to be more in line with that of the incandescent bulb would impact both the cost and the efficiency, which might put a damper on sales. If they could somehow figure out a way around that then there might be a hope for better quality LEDs. Until such time I will continue to stick with the incandescent as my choice for the light bulb with the highest quality of light.

Howdy, Shawn!

You hit the nail on the head! In a former life I was an engineering contractor for an LED lighting company. Each diode can only emit a specific frequency/wavelength of light. White light is, as you have pointed out, a spectrum - an continuous amalgamation of all kinds of frequencies/wavelengths. Think of the difference between how a vinyl record produces sound versus a CD. It would be impossible to make an LED array that exactly mimicked the white light spectrum because that spectrum contains an infinite (or if you prefer a very, very, very large) number of frequencies; it would require an "infinite" number of LEDs.

Since LED manufacturers are in the business of making money the question becomes: what is "close enough?" You can bet they are going to use the minimum amount of technology to meet the customer's specifications. Done properly a tri-color LED array can do a very good job of approximating white light; but it's still an approximation. And it's three times more complicated (expensive) than, say a single blue LED embedded in an amber casing. Incidentally that single blue LED in an amber casing is the most a common solution to LED lighting that is used for human vision.

Of course everyone has to make their choices about lighting and there is no one-size-fits-all solution, unfortunately. I hope someday I can be savvy enough to make my own candles :) But I have a few higher priorities in the mean time...    

So Bobby, knowing what you know, if all you are concerned with is the health aspect of them (not energy efficiency), would you have LED lights in your house?

The key to getting good color spectrum from LEDs is selecting the right ones and driving them at the right power level.  As you pointed out, most "white" LEDs are actually a blue LED with a yellow phosphor coating.  The phosphor absorbs some of the blue light and re-emits it up near the yellow end of the spectrum.  If you drive it at or below the designed power levels then you get a decent color spectrum output, but as you start overdriving the blue LED, you get to a point where the phosphor is saturated and can't emit any more light and then the blue starts to dominate.  The harder you push an LED the more it's output move to the blue (cool) end of the spectrum.

If you select "warm-white" LEDs and don't over drive them you get more yellow and red and much less blue.


Note even the 'warm-white" still has a substantial amount of green in the light, however, it is much closer to natural light (just less red and much less blue than natural light)

FWIW Your eyes are very sensitive to blue and green light, which makes cool white lighting useful for task lighting, reading, etc.  It takes much less power to produce the same 'apparent' brightness with cool-white LEDs than warm-white, partially because the LEDs are more efficient, but mostly because the same amount of light 'looks' brighter.

We use warm white LED in the bedrooms and the main light in the living room, with cool white task lighting in the living room, office and kitchen.

I have a question about your claim:
"Incandescent light produces a very warm light, mostly lower on the energy spectrum with a lot of orange and red in the mix. This is a good fit for eyes that have adapted for thousands of years to sitting next to a fire or lighting with candles in the evening. LEDs produce a lot of blue light, high on the energy spectrum."

Is there any scientific evidence to back up the second sentence, or is that just your personal belief?

Mike Jay wrote:So Bobby, knowing what you know, if all you are concerned with is the health aspect of them (not energy efficiency), would you have LED lights in your house?

Hello Mike! I actually had never heard of any health concerns about LEDs until Shawn's OP. I was confirming his suspicion as to the reason affordable "full" spectrum LED lights are not (and not likely to be) available in the marketplace in the near future (Peter's post is a better technical illustration than what I described). I was not intending to speak to health concerns of lighting in that post.

Since I was not aware of any health concerns, and since you asked, I do have LED lighting in my house. Energy consumption and bulb life drove me to replacing incandescent and CF bulb with LEDs. I've been doing that for almost 10 years now. Since I have no knowledge/experience relating to said health concerns I cannot speak to that topic at this point.

Gotcha, thanks!

Another thought:

The human eye is most sensitive to blue light, followed by green light, with red light a distant third place.  In fact, depending on the specific frequencies our eyes are between ten and several thousand times more sensitive to blue light than to red light.

One of the reasons that incandescent lights are so in-efficient (as a light source) is because over 90% of the light they emit is at frequencies the human eye can not see.  Because the amount of blue and green light they emit is so small (relative to red and IR) you need to use more energy and produce more total light just to get enough blue&green to see.

If you were to compare the amount of light needed to read between and Incandescent and a warm-white LED, I'm pretty confident that the Incandescent will produce more "blue" light than a warm-white LED.

Remember your eye NEEDS that blue light (and green light) in order to see well enough to read.  The fact that you are producing 9 times as much red and IR light is irrelevant to the amount of blue and green you need to read.

If we are to assume that blue light at night is actually a problem, then the warm-white LED is a better solution that the incandescent since it produces a lot of green light with relatively little blue.  So an LED reading light would actually have less blue-light than an incandescent reading light.

Peter, you questioned the science behind my thoughts on firelight and candlelight. I confess this to be pure speculation on my part. I think that it is possible, maybe even probable, but it's fair to say that's not good enough. Maybe someone has researched it but I haven't had time to find it.

I have done some further research though and here is what I found today:

"As the sun gets closer to the horizon, short wavelengths are scattered in the atmosphere and longer, redder wavelengths more easily reach the surface of the Earth. The sensitivity spectrum of melanopsin may be an adaptation to the natural solar cycle, so that ipRGCs are tuned to discriminate daylight from evening, better entraining the circadian rhythm." (source)

I think that rather than focusing on candlelight and firelight, an even better approach may be to focus on sunset. After all, for millions of years (even before the campfires) sunrise and sunset have been clocking many animals like us. I think that besides sitting in the dark or going to bed (which for me here would mean 16 hrs in the dark in the winter), the best thing we can do for our bodies is to "extend the sunset." I was surprised when doing my research that the spectrum of an incandescent bulb is nearly identical to that of a sunset.

(I acknowledge that the LED chart is a cool white LED and a warm white LED would be more charitable. This was just what I found.)


(Source)

Another interesting thing that I found in my research was a study (source) that suggested that the primary photoreceptor responsible for circadian rhythm clocking is actually separate from the photoreceptors used for vision. According to the graphic below from the study, it looks like the effect of light on melatonin production do pretty much disappear between 550 and 600nm and are gone for higher wavelengths. That looks fairly positive for warm white LED quality, at least so far as circadian rhythms are concerned.



I want to acknowledge Peter's point about needing a certain amount of blue or green light to read. I thought that was a good thought-provoking point. I think I need to sit with that a while longer.

I have spent far more time reading research papers on this than I ever thought I would and will say definitively that I will run from cool white LEDs for evening lighting. I have also read that the phosphor coating on the warm white LEDs can weaken and allow more blue light through. That concerns me, but it is hard to know how often that happens. My hope is that in the future bulbs will be designed to not do that as often.

For now I'm going to stick with mimicking sunset with incandescent lights because I feel like that is most in line with how it would have been in nature. Thanks for making me think through my position in greater detail though. It was helpful. I am much more comfortable now with where LED quality is headed. I'm hopeful that they'll keep getting better. Maybe some day I'll buy one.

Alan Booker shared some interesting thoughts with me on this matter.

The first is that one of the big problems with CFL is "flicker." Apparently it is also a concern with LEDs. While it may cause problems for people with epilepsy, it can also contribute to malaise, headaches, and impaired visual performance. Alan shared this link as a starting point.

Perhaps this one is solvable with the right design, but until then I think it's a fair concern.

Another one from Alan that I found very interesting: LED light has almost no infrared light. The sun has lots. Incandescent has lots. Apparently the infrared is rather important for all sorts of biological functions:

https://articles.mercola.com/sites/articles/archive/2016/10/23/near-infrared-led-lighting.aspx

Shawn Klassen-Koop wrote:Alan Booker shared some interesting thoughts with me on this matter.

The first is that one of the big problems with CFL is "flicker." Apparently it is also a concern with LEDs. While it may cause problems for people with epilepsy, it can also contribute to malaise, headaches, and impaired visual performance. Alan shared this link as a starting point.

Perhaps this one is solvable with the right design, but until then I think it's a fair concern.

Shawn,

Anything running on AC can flicker. This includes incandescent. If you can imagine the filament gets hot 60x a second, but it keeps glowing for a little bit after the top of the sinewave. This makes it look like a steady light to our eyes.  Because LEDs run on DC typically, they have a AC to DC converter built in. This can have AC bleeding into the DC. Most people don't notice it but of interest to me, there are people who have aberrant "timekeeping" EEG waveforms (the theory anyway) that may make them operate on a different frequency. Most of us don't get too bothered with 50 or 60 Hz, because we operate at, say, 40 Hz. But some people may not be wired the same way. After a head injury these patterns can shift, or become inconsistent. This is a frequency issue, there may also be wavelength issues which Peter has addressed quite well. There are IR diodes available (the first LEDs were IR spectrum), maybe they should be integrated for certain people. Or we could just go outside more.

One really concerning thing to me about the wavelength is the profile of HD screens (or any modern LED screen), which is really stark blue, right to the retina. This is a recipe for terrible sleep patterns. Far more than overhead lights which have a chance to be reflected.

Lets just say this is a promising area of interest. This is possibly a source for people with sensitivity to certain RF frequencies as well. I don't think they have any circuitry you or I don't, I suspect many are internally gamma inconsistent and this is just the final straw from leaking AC or induced currents. I have a friend working on this as a therapeutic intervention, and he is no idiot.  I read everything he sends me, because if there is a therapy, it could be life changing for several pathologies, including PTSD and hopefully TBI. Probably could get a coherent therapy without tech, by meditation or something, but I hear it is really hard to meditate when you are tweaked out by things that are kind of everywhere. I really hope this stuff pans out.

I think the strong correlation with sequelae of PTSD and TBI (like anxiety) with these same sensitivity syndromes often may be explainable by one root cause. Most of these people I very much doubt are "crazy", they are manifesting symptoms which we tend to mask or ignore because we think we are smart but really we know shockingly little about this stuff, especially how to repair damage.

Regarding flickering… Running the LEDs on pure DC (with a proper power supply that has big capacitors to even the output, or batteries) solves that problem. However it requires basic understanding of electricity to set it up.

Sebastian Köln wrote:Regarding flickering… Running the LEDs on pure DC (with a proper power supply that has big capacitors to even the output, or batteries) solves that problem. However it requires basic understanding of electricity to set it up.

I agree. I think that the flicker issue can be mitigated or even eliminated with the right setup. I think my concern is based more so on most people not being aware of the concerns and not taking them into consideration when setting up their system.

"This is a frequency issue, there may also be wavelength issues"
Please note that frequency and wavelength are two ways of describing the exact same thing.

Flicker frequency. Was imprcise language. Sorry.

I found this chart useful in understing lighting colors.



My desk lamp light is the last artificial light I see for the day so I have a really old small brass one with an equally old amber shade on it and put a really warm relaxing 2400K led bulb in it.  

Works great for me. :  )

Thanks Greg,

That chart seems fairly representative of the lights we had installed during renovations, the Electrician did in fact give us the correct recommendations so it’s good to see them validated i.e. we had the ‘warm white’ LEDs installed in the dining/kitchen ceilings with dimmer controls, and the ‘cool white’ LEDs installed in the bathrooms, over the kitchen benchtops, and under the outside verandahs – typical places where bright white light is needed for work and safety purposes.

I believe dimmer switches are a good addition because they lower the intensity of lighting to suit what’s coming through the windows and glass doors – cloudy versus bright sunny days. Also, they allow us to mimic dawn and dusk – I hate bright lights first and last thing in the day.

Also, auto/manual security LED floodlights (cool white) REALLY light up the yard, much better than incandescent bulbs.

As far as recycling goes, it does create an issue because there are many more components … just how conscientious recyclers are is another issue. Typically, as new technology is adopted, costs are reduced and designs usually become less resource heavy.

Anecdotally, it appears that LEDs also attract less bugs at night than fluorescent lighting. That is a major issue for us: bugs attract more frogs, more frogs attract snakes!

(I agree that incandescent bulbs provide an added heat source benefit in a cold climate. We usually don’t need that added benefit in most parts of Australia.)