The scientific reason you don't like LED bulbs—and the simple way to fix them

July 28, 2017 by Arnold J Wilkins, The Conversation
Credit: Shutterstock

There's a handy trick for reading station signs that otherwise fly past in a blur as you travel in a high-speed train. Look at one side of the window and then immediately at the other side of the window. When you change your gaze, your eyes will automatically make a rapid jerking movement, known as a saccade. If the direction of the saccade is the same as that of the train, your eyes will freeze the image for a split second, long enough to read the station name if you time things right.

Saccades are very fast movements of the eyes. Their exact speed depends on the size of the movement, but large saccades can move the eyes at the same rate as a . The image of the station name becomes visible because it is travelling at the same speed as the eye, and the images before and after the saccade are blurred and so don't interfere with the image of the sign. This shows us that our vision is still working when our eyes move rapidly during saccades.

Scientists used to think we could see no more than about 90 flashes of light a second but now we know it's more like 2,000 because the eyes move so rapidly when we change gaze from one point to another. During the eye movement, the of light creates a pattern that we can see. And this has some surprising consequences for our health thanks to the way some types of lighting can affect us. In particular, it could discourage people from using more energy-saving LED lightbulbs.

Most lighting is electric and powered by an alternating current supply, which makes the bulbs continually dim and then brighten again at a very fast rate. Unlike filament lamps and to a lesser extent fluorescent lamps, LEDs don't just dim but effectively turn on and off completely (unless the current is maintained in some way).

Health concerns

We know from earlier work on fluorescent lighting that even though the flicker is too fast to be visible, it remains a likely health hazard. In 1989, my colleagues and I compared that flickered 100 times a second with lights that appeared the same but didn't flicker. We found that office workers were half as likely on average to experience headaches under the non-flickering lights.

No similar study has yet been performed for LED lights. But because LED flickering is even more pronounced, with the light dimming by 100% rather than the roughly 35% of fluorescent lamps, there's a chance that LEDs could be even more likely to cause headaches. At best, it's likely to put some people off using LED bulbs because of the annoying, distracting effect of the flickering, which we know can be detected during saccades.

One obvious way of avoiding the flicker is to operate the lamps with a direct current so the light is constant, but this involves more expensive, shorter-lived components. Another solution is to design the lights so that the flicker can't be detected. But just how fast must the flicker be in order to be harmless?

To find out, my colleagues and I asked people to make a saccade across a flickering source of light and to report when they could see a pattern of multiple images of the light during the eye movement. When the flickered 1,000 times a second the pattern could clearly be seen. At about 3,000 per second, the images became invisible.

In contrast, some LEDs flash only 400 times per second. This flicker is still far too rapid to be seen directly, but some people can see multiple images of the lamps every time they make a saccade, which is unpleasantly distracting. The flickering of these LEDs may limit the uptake of the bulbs, just as many people dislike energy-saving fluorescent lamps.

When you buy an LED bulb, you currently have no way of telling whether or not it will flicker. But there are already standards for LEDs that would limit flicker to acceptable levels. So ensuring these are met could make a big difference to our attempt to make our homes and workplaces more energy efficient.

Explore further: Computational imaging on the electric grid

Related Stories

Computational imaging on the electric grid

July 14, 2017

Researchers at the Technion-Israel Institute of Technology and the University of Toronto have developed a technology for producing a new understanding of the nighttime landscape—from the office level to the entire city—based ...

Small birds' vision: Not so sharp but superfast

March 18, 2016

One may expect a creature that darts around its habitat to be capable of perceiving rapid changes as well. Yet birds are famed more for their good visual acuity. Joint research by Uppsala University, Stockholm University ...

Fluorescent lights keep smartphones from ten-second drifts

July 28, 2012

( -- One obvious way to reset a mobile device's clock is to check out a master clock online but Zhenjiang Li, a computer scientist and doctoral candidate at Hong Kong University of Science and Technology, and colleagues ...

From the desk lamp to the desktop?

March 9, 2010

In the future, getting a broadband connection might be as simple as flipping on a light switch. In fact, according to a group of researchers from Germany, the light coming from the lamps in your home could one day encode ...

Recommended for you

EPA adviser is promoting harmful ideas, scientists say

March 22, 2019

The Trump administration's reliance on industry-funded environmental specialists is again coming under fire, this time by researchers who say that Louis Anthony "Tony" Cox Jr., who leads a key Environmental Protection Agency ...


Adjust slider to filter visible comments by rank

Display comments: newest first

not rated yet Jul 28, 2017
I have LED lights in my ensuite and most mornings I start in there, then head to a darker stairway to downstairs. While walking down the stairs I frequently have a strobe effect of flashing light. I don't see an image of the light bulb used, just flashing white light.
1 / 5 (1) Jul 29, 2017
I have LED lights in my ensuite and most mornings I start in there, then head to a darker stairway to downstairs. While walking down the stairs I frequently have a strobe effect of flashing light. I don't see an image of the light bulb used, just flashing white light.

I bought one of those LED imitations of incandecent bulbs with the yellow LED strands for filaments. It was "reasonably" priced, only $5 so I thought I'd give it a go.

In a dark room, shaking my hand underneath it I can count 10 distinct fingers.

The problem appears to be that the Chinese engineers thought they could use one strand for one half-cycle and another strand for the second, to push the flickering up to 100 or 120 Hz depending on market area, but they didn't realize that the strands would be pointing in different directions, so the whole bulb is effectively blinking from left to right and anything that moves in the light gets a strong stroboscope effect.
1 / 5 (1) Jul 29, 2017
The irony is that long fluorescent tubes with electronic ballasts - old tech - is more energy efficient than the average LED bulb.

But the market is too competitive and the prices are too low, so the industry is drumming up LEDs instead because they have higher profit margins and as it's a fairly unknown novelty, so even complete crap products sell to the early adopters and other blue-eyed acolytes.

After all, who here has an integrating sphere to check the actual lumen output of a diode at its rated current and voltage? Nobody, and people can't easily tell the difference by eye because the blinking and the differences in color spectrum fool the eye, so you can sell just about anything and claim everything about it - and nobody is the wiser.
not rated yet Jul 30, 2017
Don't blame Chinese engineers on the dual half cycle scheme of filament style bulbs. Blame the consumers who wants something incredibly cheap and who like the look of a glowing filament.

And do some research!!! It's easy to get an LED bulb that is over 100 lumens/Watt. That's the maximum efficacy you can get from a fluorescent tube and you're luck to hit 80 lumens/Watt once a fluorescent tube is installed in a fixture.

The margins for LED bulbs are about are far worse than traditional incandescents. Everyone has been cutting margins (they think they'll make it up on volume) to gain market share. The truth is that most Chinese companies are losing money on LED replacement bulbs.

In the US you have to have a Lighting Facts label on the packaging of anything with an Edison base. You don't get to make up the numbers, you have to submit a testing report from an independent and certified lab!
1 / 5 (1) Jul 31, 2017
Blame the consumers who wants something incredibly cheap and who like the look of a glowing filament.

That's beside the point. These lights are being manufactured extremely cheaply, shoddily, and sold at a very high markup because the consumers aren't aware what they're getting.

It's easy to get an LED bulb that is over 100 lumens/Watt

Two points.

1) Those are usually based on false advertisement. The diode manufacturers spec the chips at a certain voltage and current, which results in a certain luminous output - but the industry standard is to measure this at 25 C junction temperature. In the real lighbulb, the temperature is different and the luminous efficacy is less.

2) the LED lamps that do achieve over 100 lm/W with good color index, like the Philips L-prize, are based on secondary emission from a phosphor target, very similiar to the fluorescent tube. They also cost $25 a pop. The cheaper ones achieve high lm/W at the expense of poor spectrum.

1 / 5 (1) Jul 31, 2017
In the US you have to have a Lighting Facts label on the packaging of anything with an Edison base. You don't get to make up the numbers, you have to submit a testing report

You're confusing two different labels.

The DOE label is voluntary, requires standard independent testing. This is used by/for industrial LED products.

The FTC label is mandatory, but does not need an approval process to print on the package. They take you on trust that you have done it according to their testing requirements, which are generally the same as for the DOE label.


This testing provides a snapshot of performance under specified operating conditions (...) It does not address lifetime ratings, changing performance over time (e.g., lumen maintenance), or LED case temperature.

There's considerable room for cheating.
1 / 5 (1) Jul 31, 2017
Does the FTC verify performance claims?

The FTC label does not require test procedures to verify the stated performance claims. The FTC encourages stakeholders to refer to the DOE Lighting Facts program for verified performance values for SSL.

They're specifying which testing procedures you should use, but then don't check upon whether you've actually done it properly.

Besides, the Edison bulb imitation LEDs may fall under the category of "retrofit products" which appear to be exempt from the labeling requirements.
not rated yet Jul 31, 2017
You should really learn something about LED lighting. For someone who claims to know all the dirty secrets of the industry, you don't seem to realize that it is virtually impossible to buy a (white) LED replacement bulb that doesn't use phosphor conversion. The Philips L-Prize bulb was an absolute economic failure because it was simply too expensive. It was the first high volume LED light bulb to exceed 100 lumens/Watt, but at >$30 it sold to a handful of LED enthusiasts and almost no one else.

The only RGB white LED bulbs on the market are the Philips "Hue" products. They are incredibly expensive and the color change feature is more of a novelty and not user friendly.
not rated yet Jul 31, 2017
I also stated that Edison base products need a Lighting Facts label. Which by default would have been the FTC Lighting Facts label and not the Energy Star Lighting Facts label.

Both require a test measurement made by a NVLAP accredited testing facility for the relevant test (IES LM-79 for LED based lighting products). NVLAP is a National Voluntary Laboratory Accreditation Program which is a NIST program to accredit unbiased labs according to the ISO 17025 standard.

You can't do the test in your garage and use that to make your label.

There isn't room for cheating. Although LED manufacturers may test their LEDs at 25 C (some test at 85 C), LM-79 requires measurements to be made at thermal stability which is defined as <0.5% change in luminous flux over 2 consecutive 15 minute intervals.

Do your homework on an industry before you slam it. Spreading rumors and false information does no good unless you are trying to troll.
not rated yet Jul 31, 2017
As for your first reply, something can't be sold at a high mark up if the Chinese manufacturers are going bankrupt. Everyone shaved their margins to the bone in an effort to gain market share.

World-wide LED lighting companies have been going out of business and there has been a lot of market consolidation because no one is making money.

Why do you think that Philips and Osram sold their LED divisions? They didn't do it because their profits were too high.

Big box stores like Lowes and Home Depot make their margins, but they have margin requirements for everything they sell. It's not unique to light bulbs and they may be the only people making money off LED light bulbs.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.