Why do you think that the hysteria around the blue peak in the spectrum of white LEDs is false?

Go ahead show any *evidence based* confirmation that blue peak ever did any harm to anyone's health out of *billions* of users, right?
Not that yellow press blabber...

To collect evidence of damage that takes years to accumulate, and then correlate it back to exposure to LED lighting, out of many additional factors that may lead to the same results (increased screen time on the smartphone, reduced outdoor time, general stress, changing environmental factors, ...) is not an easy task, if possible at all

However, the evidence of absence of damage, on which all "LEDs are safe" claims are based, are literally "we have put rats under grow lights for 48h and they did not complain"

Well, the Philips Ultra Definition bulb I used in the stairwell to my basement just had a driver failure; it is very dim (not even on a dimmer switch) and flickers quite a lot. However, it has a smooth flicker, indicating the presence of a smoothing capacitor. Still appalling for a bulb that has been there for at most a month, and not even running the whole time too!

To collect evidence of damage that takes years to accumulate, and then correlate it back to exposure to LED lighting, out of many additional factors that may lead to the same results (increased screen time on the smartphone, reduced outdoor time, general stress, changing environmental factors, ...) is not an easy task, if possible at all

However, the evidence of absence of damage, on which all "LEDs are safe" claims are based, are literally "we have put rats under grow lights for 48h and they did not complain"

White LEDs for general lighting are already with us for quite long time, somewhat 15+ years. If there are any health risks, these would definitely show up, but yet just none... And seriously, white LEDs are not the exclusive light source to have prominent blue peaks, see mercury lamps, all the fluorescents. These are known for 80+ years at least, but again no problems was discovered yet...

The world is full of useful idiots, of course, who invent incessant conspiracy theories, sure to someone's profit. Killer LEDs is one of these Do you remember 'killer CFLs' haha
 

Here is a quick comparison of some bread-and-butter, not excellent, but definitely not bad 830 color LED lamp, a kind that you expect for home lighting VS a  circline, good old 530 halo color, apparently made by Panasonic. You still dare to say the LED is a killer while the fluorescent is OK? Fluorescent lamp has blue (and violet!) spikes of comparable intensity, and 530 color is known and is in use at least from 1950s! And fluorescent's lamp blue peak is even more energetic at 437nm.

I used spectrum plots out of Philips datasheets instead of your pictures, for a few reasons :

1. I want to try to make some measurements out of the pictures. The Philips pictures are of better quality and aligned (not taken with a camera from the screen of another device)

2. I expect that the Philips measurements were done with a better precision tool, especially regarding the blue peak of the mercury discharge which is a very narrow line. Ideally this should not matter that much (your HP320 makes it appear wider, but ideally would also average the intensity by the same amount, so the integral would be the same), but i don't know how close the instrument is to expectations

3. The concern of blue light is higher with the higher temperature lamps. While i still dont claim to be sure that 2700K LEDs are safe or that 6500K are killer, 4000K is more representative of the range where i think there may be something to compare

4. I chose both lamps of 4000K



Now, i try to do actual measurements out of the pictures, using paint program

First, all spectrum plots (of the lamps and of the functions that define the test criteria) are converted to black and white, and stretched in X direction to be in the same scale (100px = 100nm) and range (400..800)

What i calculate is the part of the spectrum of each lamp that matches each function, ie :

Integral (from 400 to 800) of SpectrumLamp * TestFunction

First the test function is converted to a grayscale image :

 - The CIE just by the paint program

 - the hazard i had to do manually because it is on a log scale. i did only few steps, but they are still the correct values (e.g. 100%=255, 80%=204, 10%=25 etc)

Then the pictures are imposed on each other

Integral done by resizing the picture to 1x1px (letting the paint program average the entire image) and testing the color of the pixel (from 0 to 255)

Final content of blue light determined by part of the blue light of interest (acc to hazard function) out of the overall visible light (acc to CIE function)



Results :

The LED light have approx 18% more blue in it for the same color and CRI

This may not be a lot (and i can't claim anything at all about the precision of my paint-experiment), but i also don't know where the threshold for health risk actually is



Considerations :

The S cone cells (which sense the blue light) are only 2% (!) of the total population of cone cells in the eye, but deliver 1/3 of all the information we see

They are also the most sensitive (take the longest to recover after over exposure etc)

They are also the first to degrade just with age (including in people who never used LED lighting, before it even existed)

This is a reason to think that those cells are already running at their limits by nature, therefore even not so significant increase might cross the limit



Finally, our body does have means of giving an advance warning about some things it does not like. Not all of them (e.g. some poisons, ionizing radiation), but bad lighting in the visible light spectrum is covered for the most part

Personally i do feel higher load on the eyes when using LED light, especially >=4000K. And this is while i use myself Fluorescent light of /765, /865 and /840 a lot and never had an issue. (At home there are all of them and some /827 too). My mom's reaction to LED is equally bad, and yet no problem with /840 PL's

Some people do report similar reaction to Fluorescents of different types, and although there may be unrelated factors involved (bad lighting distribution in the room, flicker, etc) there is no reason to assume that all of them are wrong. So yes, for those people CFLs may be a "killer", while for me and you they are not

Well a lot of pseudoscience here, but at first it is extremely unprofessional to make comparison between two spectrums when the first one is expressed in some unknown units of 'intensity percents' while the second is expressed in some strange units of microwatts/5nm/ lm (!) Or was it per 1 m^2 with ^2 dropped by an unscrupulous editor?

At least both of my measurements were done with the same instrument which is *calibrated* in mW/m^2/nm. Looks like you dropped my result because it will give you an unfavorable conclusion, and so jumped to 4K lamps data. Not fair! I can attach  .csv files with data series if you are interested to do the calculations *properly*.

No one really cares for your (and your mom's) autosuggestion regarding eye strain with different light sources. That even should never been mentioned if you are going to look any serious. Scientists use a series of methods devised to get rid of autosuggestion. Blind and even double blind tests, exactly.

BTW, what is the source of that 'hazard function'?

And in the end, some common sense tells that if two lamps have the same measured color temperature, it roughly means the same proportion of energies that are landing in the bands of red, blue and green receptors.

Also what is the source of "but deliver 1/3 of all the information we see" - rather contrary it is known that the eye has poor spatial resolution in blue. That is known for long, used by many image compression algorithms (starting with NTSC TV yeah!) and this is why that idiotic blue 7-segment led displays have bad readability)

@RRK: I would recommend you to see this video of mine to see how much blue light 3000K LEDs contains. And yes: Insects being attracted to blue light as well.

Why do you think it is a blue *peak* that attracts insects? Even regular incandescents attract insects, this is why specialized yellow lamps exist.

It of course may be new to you, but a regular 2700K incancescent lamp light output does contain some blue too. And some violet too. And even some minor UVA!

Read here for more information: https://en.wikipedia.org/wiki/Light-emitting_diode#Health_and_safety
If it is false, you can correct it, using the correct cites.

So what?
It is already all honest.

"One study showed no evidence of a risk in normal use at domestic illuminance"

Don't worry, be happy ))