I think there is a wide range of Red spectrum that has biological effects
I do not recommend pointing red LEDs at your eyeballs
I have emailed the author of the paper, and asked if he wouldn’t mind popping in here to help us all with this - here’s hoping! 
This looks clearly like a 620nm, because of the orange-yellow center
Deeper reds like 660nm have more magenta there than orange-yellow.
That is a bit of a rule of thumb do differentiate between reds with the eyes.
Highly interesting.
I wonder what are the people that are heavily into Red LED therapy have to say about this?
I mean currently red light is considered also a metabolic booster, so in theory that should do similar things as this Green LED does, here I mean like faster wound healing.
Ok, I just had a very kind reply via email from Glen (the publisher of the paper) at UCL.
He tried to post the below on here, but was having trouble for one reason or another.
Here is the email, I hope it helps you guys - I’d like one too lol!
Hi ,
Just tried to post this but don’t think it worked
Hi I’m Glen who published the 670nm paper and human vision. Academic at University College London. G0OSE asked me to join in. Few comments in general.
We have been working for >8 years on interactions between light and biology. We first improved vision in flies and bees then in mice with 670nm. There is a sound scientific rational. Our human work is a simple extension, same metrics, same results. You can see my labs other publication in Pubmed under Jeffery G.
In flies and bees long wavelengths also improve metabolism, memory and mobility in aged animals. Its based on the notion that mitochondria in cells provide energy for cell function that declines with age. Mitochondria absorb these wavelengths and it improves their energy output and cell function. It is simply like recharging a battery. A number of major clinical trials are running with the aim of improving mitochondrial function in aged disease.
Wavelengths are 670nm- approx 1000nm. energies were around 40mW/cm2 to start with. This was filtered, but we now have as good effects with much much lower power output. There is no real dose response curve - the effect is more switch like. 3 mins as good as 15. >1h and the effects are reduced.
The process is often termed photobiomodulation. There is a lot of literature out there, look for names like Mitrofanis J. Kern T., Eells J., etc,…
A patient came to me with a single 670nm LED and a battery. she joined them with her fingers and asked me if it would work. I tested it and said yes. Nothing magical. Don’t know of anyone making these devices commercially, but bet they will soon
I can’t always respond, but will try
Best wishes
Glen
Hi I’m Glen who published the 670nm paper and human vision. Academic at University College London. G0OSE asked me to join in. Few comments in general.
We have been working for >8 years on interactions between light and biology. We first improved vision in flies and bees then in mice with 670nm. There is a sound scientific rational. Our human work is a simple extension, same metrics, same results. You can see my labs other publication in Pubmed under Jeffery G.
In flies and bees long wavelengths also improve metabolism, memory and mobility in aged animals. Its based on the notion that mitochondria in cells provide energy for cell function that declines with age. Mitochondria absorb these wavelengths and it improves their energy output and cell function. It is simply like recharging a battery. A number of major clinical trials are running with the aim of improving mitochondrial function in aged disease.
Wavelengths are 670nm- approx 1000nm. energies were around 40mW/cm2 to start with. This was filtered, but we now have as good effects with much much lower power output. There is no real dose response curve - the effect is more switch like. 3 mins as good as 15. >1h and the effects are reduced.
The process is often termed photobiomodulation. There is a lot of literature out there, look for names like Mitrofanis J. Kern T., Eells J., etc,…
A patient came to me with a single 670nm LED and a battery. she joined them with her fingers and asked me if it would work. I tested it and said yes. Nothing magical. Don’t know of anyone making these devices commercially, but bet they will soon
I can’t always respond, but will try
Bet wishes
Glen
Yay! Welcome to the BLF Glen, so kind of you to join us here!
Your input is VERY welcome!
Where are people freely getting these specific 670nm LEDs since that is not a commonly available wavelength for red LEDs.
620nm and 660nm are the basic ones from 99% of the red LEDs
Some how I suppose my posting worked!
Best
Glen
WOW that did weird things to my eyes when I went back to a white page lol! psychedelic! presumably best viewed in the dark? I wonder is it accurate, is it safe?
Fantastic solution if it’s accurate, but I suppose that will depend on the monitor.
Correct! You clearly know what you are looking at.
I dont think we need to be so specific
here are some options:
thank you moderator007
Thank you Sir!
very grateful for your information… lots of questions… not all may be relevant… just wondering out loud.
how many lumens do you think is not too strong for the eyes?
I have 4 levels to choose from on this particular light
0.1, 2, 26, 105 lumens
Do you know if Green light also charges mitochondria?
Do you know if other LED colors, also charge mitochondria?
Do you know if Sunlight also charges mitochondria?
Again why 670nm and not 660nm which is so commonly available from no-name LED manufacturer to well known LED manufacturers.
If you look into the graphs of most deep red LED their peak is 660nm not 670nm. While many are rated 650nm to 670nm does not mean their peak is 670nm.
Example here, specifications and peak are not identical that is why most manufacturers say 660nm and not 670nm
https://eu.mouser.com/datasheet/2/602/DS236-luxeon-2835-color-line-datasheet-1596094.pdf
Yeah, that is what I am trying to say, I do not think we need to be that specific as 670nm instead of 660nm when that is where the peak is of 99% of deep red LEDs.
Hard to start posting all datasheets now I looked over but I did post one to give as an example.
The specific wavelength doesn’t seem to matter, notice the broad range that Glen identified.
imo 670 is not specifically required… it just happened to be part of the range used.
As you said, LEDs are not actually single nm output… I dont think it is necessary to only use 670nm
there is a tendency to latch on to one particular bit of info (the number 670 in this case), and then assume that is the key ingredient… I do not believe that is necessary.
Bingo!
Agree
now, who is ready to inform us whether 620nm works fine too?
Yes, they all have a broad spectrum and we are only talking peak here. There is so much we do not know. Vast amount of work on 670nm. Shorter wavelengths are generally less effective. But we have not mapped out others that well. I have used 810nm which was fine. But then you can’t see it so there is the insecurity of about if it is working or not.
670nm LEDs are a little difficult to obtain and in my experience been of varied quality.
Maybe you can test some 660nm since most manufactured Deep Red LEDs peak at this wavelength. And there are of course the typical 620nm peak for regular Red LEDs.
i would assume that the 40mW/cm^2 is the radiant flux hitting the surface of the eye, so for a 5mm pupil that would be about 8mW of radiation power into the eye. This would be over all electromagnetic frequencies, but for semi-monochromatic one could assume it was mostly at the frequency of interest.
Lumens is a photometric unit for the luminous flux, or the total visible light emitted by the source. This is related to the spectral response of the human eye to visible light wavelengths.
So in general terms there is no direct conversion factor of lumens to irradiance, one is the power emitted into a sphere, the other is the power hitting the target.
okay so i’m gonna go outside and stare at the sun like Barkuti and see what happens…
yes, Lux is different than Lumens
but, if I put a red light directly on the target, Lumens and Lux are the same.
can someone help me figure out how many lumens from an AA Eneloop, produces
40mW/cm^2 on a target that is touching the bezel?
or let me ask that in another way
at what distance would a 2 lumen source, produce 40mW/cm^2?
and fwiw, there are mitochondria in blood
so maybe aiming a red LED at my heart will help treat the whole body against inflammation? Or aim the light at the femoral artery, to reduce foot swelling?
lots of possibilities…