Optical Society publishes study on the perception of white LED light

Age combined with variance in short-wavelength energy in white SSL result in varying perception of phosphor-converted LED white-light sources.

Maury Wright
Oct 22nd, 2019

The Optical Society (OSA) has published research in its Optics Express journal that suggests human perception of phosphor-converted, white-light sources varies with the age of the person viewing the light.

Presumably, the short-wavelength energy in the blue or violet spectral region is the culprit, although it appears the field of view is also an issue — a phenomenon we have reported on previously. The intent of the research is establishing a knowledge base that solid-state lighting (SSL) product developers can leverage to deliver consistent chromaticity perception to everyone who views an LED-based product, ranging from general illumination lamps and luminaires to backlit displays.

The research was performed by Soraa scientists and engineers, including research team leader Aurelien David. “Our study revealed that the amount of short-wavelength light a source emits together with the viewer’s age may lead to very different perceptions of a white light,” said David. “LED light sources exacerbate this effect, because their emission often features large peaks of blue or violet radiation.”

The research relied primarily on Soraa’s blue-free (BF) LEDs that integrate a violet pump and a phosphor formulation that produces virtually no energy in the blue spectral region. We published an article on that technology a couple of years back and its intended application in circadian or human-centric lighting. Upon seeing the press release on this study, we assumed there was some circadian aspect to the research. But it appears that BF LEDs were simply a good match for the work because of the prevalence of the extremely-short-wavelength violet spectral energy.

The researchers had volunteers view several near-white LED sources with different spectral power distributions (SPDs) and were asked to compare those sources with a reference white source. “According to standard colorimetry methods, all users should have the same perception for a certain LED source,” said David. “However, we found large variations between viewers. Some thought a given source appeared very pink, whereas others thought the same source appeared very green.”

The press release we linked in the lead paragraph suggests primarily that age is the reason for the differences in perception. But reading through the more detailed article, it appears that there are multiple factors at work.

The research studied LEDs that were binned primarily on the 1931 CIE color space work that is still widely used in the SSL world. We published a seminal, four-part series on color science early in this decade that details the science behind the CIE work. It’s a long read, but start with part one and make it through part four and you will learn a lot. Part one introduces the concept of metamers where two light sources with different SPDs can appear to humans to be the same color. And the series explained the concept of color-matching functions based on the science.

But that 1931 work was based on a field of view of only 2°, meaning that only the cones in the center of the eye were considered. We had a contributed article a few years back that detailed this phenomenon. There are few to no blue cones that are located within that field of view in the eye.

There was later work that identified that a 10° field of view would more accurately characterize what a human perceives in terms of color. Osram Opto Semiconductors, in fact, had a program promoting 10° binning of LEDs as a better approach than the widespread 2° binning in use. Osram engineer Alexander Wilm won our LEDs Magazine Illumineer of the Year Sapphire Award in 2017 for that work. Note that if you read this immediately after our posting, there may still be time to submit a Sapphire entry.

The OSA article discussed a new color-matching function called CIEPO06 based on a 6–10° field of view that could help with consistency of perception if applied in LED binning. Still, the study also finds a solid link between age and color perception as well. “The perceived tint of these unusual LED sources is heavily influenced by the viewer’s sensitivity in the violet range, which is strongly age-dependent,” explained David. “Although others have observed similar variations in white-LED perception, the effect was especially easy to see because these special LEDs exacerbate it.”

Original source:

[Clemence]

that is surprising, thanks for the info

I wonder,
what was the age of the ones that saw pink, and
what was the age of the ones that saw green…

was the perception using identical ambient light adaptation?
and using the same reference images before seeing pink or green?

No surprise to longtime readers here: Flyboy reported a dramatic color shift after a cataract operation. I saw the same thing after I had one.

That thread has this link to show how the lens of the human eye becomes more yellow-orange with age

And there’s much more out there, if you search:
https://www.google.com/search?q=eye+lens+yellow+age&client=firefox-b-1-d&source=lnms&tbm=isch&sa=X&ved=0ahUKEwi28ZmqjeXlAhXMHjQIHcGDDzUQ_AUIEigC&biw=1409&bih=795#imgrc=hFNQOj7qlOl_aM

PS —- anecdotally at least, this color shift also explains why you’ll see elderly women with blue hair.

Gray hair looks yellowish through older lenses, so the people making hair products add optical brigheners to them that make the hair brighter in the blue end of the spectrum, overcoming the yellow shift that happens with age. So to the elderly customer it looks —- ta daaa! —- bright white.

https://www.google.com/search?client=firefox-b-1-d&q=optical+brightener+gray+hair

Interesting. Do you know how much yellowing can occur before the doctors say that person has cataracts?

Both of my eyes see tints substantially differently. Ive also noticed a dramatic change in my ability to see at night. I’m 47 and a long term welder.

Im old, cataracts, only see out one eye
I do really well on color tests though

does anyone here, see pink where I see green?

Looking at the data, it’s a bit messy.

However, it’s worth pointing out that the study intentionally did not use normal LED light sources. It showed people some broad-spectrum lights as a reference, then asked participants to quantify the tint of some intentionally strange LEDs. The LEDs were strange because they had the blue part of the spectrum cut out, and used violet for the shorter wavelengths instead. They were specifically testing people’s ability to perceive violet light and how that ability changes with age.

The result was that people seem to lose the ability to see violet as they get older. On average, what people rated as “neutral” changed from about +1 mduv at ~20 years old to about –3 mduv at ~60 years old. But there was quite a bit of individual variation, with some of the younger people preferring as low (rosy) as –7 mduv, and some of the older people preferring as high as ~0 mduv. However, virtually everyone agreed that +4 mduv was quite green.

Also, the test used 2700K light sources… which might make things a little weird. I’m not sure why they didn’t use something more neutral like 4000K or 4500K.

The research performed by a lighting company has of course always the danger of not being independent. But actually I worry more about the quality of the experiments, it takes good researchers to perform experiments with human subjects in a way that you can draw objective conclusions.

My left eye, in certain situations which I haven’t completely mapped, is cooler/bluer than my right. The color scheme of BLF actually is a perfect palette to emphasize this. I noticed it as a kid in elementary school walking down the hall. I noticed that the large globe lights hanging from the ceiling looked different colors when I closed one eye. I don’t have corrective lenses is any kind and as of now (knock on wood) have perfect vision. I’ve scored average to high on color tests as well.

I suppose the point I’m making is with or without this study, I believe that eyes are not created equal. They surely change with age. It makes sense too intuitively that like hearing, the frequency range that we can perceive narrows and that the shortest wavelengths are the first to go. Let’s not forget about color blindness and tetrachromacy…

> Joshk

All of it. The yellowing is a change in the lens color, and can be evenly distributed as I understand.
If it’s uneven and interferes with vision, I guess it’s called a cataract.
When I had a cataract diagnosed the opthalmologist explained it as a developing defect in the lens structure, which is built up in layers; he said areas of the crystallin protein of which the lens is built can fail and separate causing an optical defect.

Here’s a page of explanations: https://duckduckgo.com/?q=crystallin+lens+yellowing finds

TL;dr: “Life is like stepping into a boat that is about to sail out to sea and sink.”
—Shunryu Suzuki Roshi

older people have yellower lenses,
they perceive tint to be further above the duv

to compensate, they like pinker tints
to bring the perceived tint closer to the BBL

My eyes are old, and my lenses yellowing
I am highly sensitized to recognizing green tinted LEDs
and my favorite LED is very Pink, sw45k

people who use a flashlight when their brain is white balanced to daylight, will find daylight duv does not appear green, even though it is above the Incandescent BBL

people who use a flashlight when their brain is white balanced to incandescent, will find green tint (above the bbl, like daylight), to be very obvious, and will prefer LEDs whose tint is not above the BBL (positive duv as in daylight)

examples of LEDs with tint above the BBL, LH351d 5000k, SST-20 4000k
these are useful when the brain is daylight adapted
will seem green when the operators brain is incandescent white balanced

examples of LEDs with tint below the BBL, N219b, especially sw40 9080 and sw45 9080
more useful when incandescent white balanced
will seem pink when the operator’s brain is daylight white balanced

the effect of white balance on the perception of color and tint

This is why my EDC is usually right on the BBL (on most used level which in this case is just after the first flash in Anduril when 7x7135 start to kick in). It appears only slightly pink in daylight, and only slightly green at night. Right now that’s a shaved dome SST20 4K in an FW1A

good compromises
in my case I choose slightly pinker in daylight, so there is no green at night

I try to match my choice of LED to the CCT of the ambient light
in actual use I find I like my flashlight to be slightly cooler (500-1000k)

if I had to pick just one LED, it would be 4500k

this is not my photo, but I like the same LED options
(and all are High CRI, the only way to fly)

they dont all get the same amount of use, but its a real luxury to have a warm LED available on the nightstand

here are my most frequently used lights and LEDs atm

given my strong love of pink, the outer two lights, though excellent in their own right, are seldom used

Nice collection _slider!

5000K is my max generally and 3000k min for 95% of use cases for me. At night, 4500K seems shockingly cool to me at times. I almost never find my shaved-dome FB4 (which checks in @ 4100K) to look out of place in almost any scenario. At work under florescent or LEDretrofit lighting it often matches perfectly. At night, the great R9 really compensates entirely for the tint when not looking at a white wall, and it never looks cooler than “pure white”. Morning or evening daylight- great there. Overcast daylight makes it’s noticeably warmer, but again not objectionably so. I’m just very happy with it! (clearly haha)

I’ve also really come to like LH351D in 5000k and also the old KD batch of 3K SST20s. A 4350K 219B is the last one that makes it into the rotation.

that happens to me too, I was suprised, until I finally started to grasp the effect of white balance of my brain

I recently moved into a house that has very warm LED lights… I would guess as low as 2200k… I do not love them. I miss the R9 CRI and full spectrum of Incandescent.

Ive been looking at some of maukkas LED light reviews, and would like to try a couple of the better ones

this review looks good

this review also looks good

this review is bad, due to flicker, I would not buy

At least now when people complain about tints and color rendering, we know to ask “and how old are you?”

it is not really age that make your eyes yellow, but contaminants in blood, lymphatic liquid, which does accumulate with age, but depending on state of health, and how effective your body is at detoxing itself. it can be very different for 2 people of same age.

I guess it makes sense. My favorite EDC right now is a Emisar D4v2 with the 4000k SST-20. I was a little hesitant buying it because I heard over and over how green those LEDs are. I’ve had some other lights (SureFires and Olights specifically) that were so green I didn’t want to use them, but I already have XPL-Hi lights so I went ahead and tried it. To my surprise, I can’t detect any green whatsoever at any brightness level. It’s just a beautiful creamy tint from moonlight to turbo, and the others sit in the drawer. It goes everywhere except on the rare occasions I think I’m going to want something with more throw.

Same perception here. I’m impressed by the 4000K SST-20, just not in the Emisar (3000K there since I heard the same thing as you did), but in a Fitorch P25. The more power it gets, the nicer the tint is. Ordered the new Amutorch E3S with the same LEDs.

My eyes are 52yo. I instantly saw the green tint of the Samsung LH351D 5000K (Wuben E05-I), later learned it was above BBL and the impression of it being greenish are widely shared. Interestingly, I disliked it in the Wuben, but used this LED from a different source in another light. Still green, but somehow still nice. The Samsung appears to be very special.

I don’t think I see colors much different from other people. Yet. But my eyes are different. Left one has a colder perception than the right eye. I always found this funny, especially as kid. I often played with the different impressions by blinking with the eyes :smiley: .