Interesting discussion between Dileep George and Lex Fridman about vision and visual cortex in the brain.
Short summary:
“The eye sees only what the mind is prepared to comprehend.”
- Henri Bergson
I could not see any datasheet to show a clear 670nm as a dominant wavelength for a red LED, while most they do have a bit of 670nm or they can be specified as range rather than a dominant they are only residual at 670nm and 660nm is the dominant wavelength anyway.
Would you share some of these dominant 670nm red LEDs datasheets that you have tested?
Since I see a massive emphasis on 670nm like it would be a standard red LED and it is not, or what exactly is the point of using this exact number of 670nm unless we can actually have one to buy.
Research in the field that was started>15 years ago used 670nm. Partly this was due to a device that I think is not longer on the marked called a Warp that most labs used. There was no general interest in the public having access to 670nm as all of the research was animal based with only about 20 labs working on it. Since then it has expanded with a demand for 670nm LEDS - Try Epigap they normally stock them
https://www.epigap-optronic.de/en/led.html
All LEDs will have a dominant wavelength and a normally distributed wavelength spread. As long as 670nm is within the half power bandwidth it will probably have some effect. Sorry I’ve not retained data sheets. We check and throw out what is incorrect. This varies with batches even from the same supplier.
Best
Glen
Hank Wang from https://intl-outdoor.com/ just told me that his SST-20-DR leds are 670nm for now. I don’t know if that number was measured or if it is the manufacturer bin.
SST-20-DR is advertised as a 660nm led, but according to the data sheet it has 6 chromaticity bins, between 640nm and 670nm.
https://download.luminus.com/datasheets/PDS-003040-Rev-04-SST-20-DR.pdf
You can order a custom Emisar or Noctigon flashlight from him with SST-20-DR leds if you email him. Emisar D4V2 and Emisar D4SV2 can both be made with these leds. You can find his email on the website. I purchased both and I like the larger D4SV2 more, because it has a more throwy beam.
EDIT: I updated this post because those SST-20-DR leds were actually D5 chromaticity bin, 660nm-665nm.
I’ve had severe myalgic encephalitis for the last 10 years and a couple of years ago my mitochondrial function was measured to be well below normal which, as you can imagine, causes a myriad of issues.
My knowledge is a mix of scientific fact and non-scientific self-observation so difficult to quantify, but before i knew about the mitochondrial dysfunction i looked at infra-red light to see it’s effect on certain issues and discovered that looking at an infra-red heater panel for a few minutes ‘relaxed’ my eyes and benefited my eyesight and light sensitivity. (Caveat: I was far enough away from the panel to not feel any change in temperature on my eyes, don’t want to burn them!)
Obviously the research stated in this thread could go some way to explaining that but the reason i’m saying this is because although this is a torch… ok, flashlight forum an infrared panel might be more conducive to this health application given that it is a far less focused source of light.
Although sticking a DR/IR LED in a suitable host is far more in keeping with the nature of this forum 
It is also worth me stating that the light from the 660nm XP-E2 in the C01R relaxes my eyes whereas the 620nm XP-E and XM-L colour that i also have fatigue them, i actually prefer white LEDs to 620nm LEDs.
And, if it is of benefit to Glen i have a new C01R i can send which is already in England.
I tend to agree with you re the panel particularly for systemic mitochondrial disease. This is a bit off fields for me, but its where some good researchers are going who are working on Parkinson’s etc, which has a mitochondrial basis. The other bit that adds to the potential strength of the approach is that we know mitochondria in different parts of the body talk to one another and can communicate about health status and shift as a consequence. After all, mitochondria regulate ageing and generally the organs in our bodies age at similar rates, although the retina leads.
However, for many of us it has been a long journey to get people/scientific community to believe in what we are doing. To then go and tell them to direct the light on a part of the body that is not the focus of pathology is problematic. But there is a very solid scientist in the USA who has improved retinal function in rodents by focusing light on the back of the animal. Not quite as good as focusing on the retina, but impressive.
It’s a fast moving field and while I can put my hand on my heart and say this is all working and showing improvements, I and others are very aware that there is a lot we do not know.
Best
Glen
Science!
I wanted to try this “method”,so I searched around the Internet and fount some multi led deep red lights on Amazon,named ABI deep red grow lights,with different power.They say that they are of 660nm.
Being in a hurry,as a friend was to come in Athens,I told her to buy two of them,one with 12W power,and the other with 24W,as I didn’t know exactly the details.She gave me these today,and I tried one at a time.
The result is that the big one,is way strong,so it is not possible to look at this,even from a far distance.The 12W one,is more tolerable,but not very.I tried from about 1.5m far,but the light is very strong.So,I will try this from a 3m distance and will see.If I have a result,I right about.
These lights have a fan inside,because of the heat which the leds produce.
for those interested in red LEDs to install themselves
Cree-XP-E2-Photo-Red-660nm
A host I like:
Tool AAA
for those that just want to pay and play,
here is a Photo Red in a twisty host
Way too much power! and you need to use a diffuser such as a piece of waxed paper or optical film. Don’t look at a bare LED.
Vision Advisory
WARNING: Do not look at an exposed lamp in operation. Eye injury can result.
Thank you very much,appreciate. 
Well, you bought a mortar and a howitzer to kill flies. :-D
A very small amount of light is enough! You should be able to look directly at the light source without discomfort. This means the light source's emission power needs to be tuned to obtain a certain amount of power density per surface area. Read this comment of Glen Jeffery, even 40mW/cm² is way more than really needed. A proper optic or filter to evenly distribute the light is also quite advisable.
Now that I think of it, 40mW/cm² sounds like a lot. Let me also say that an aspheric/plano-convex lens just atop of the emitter is, in my opinion, suitable for this ensuring a superbly uniform blanket of irradiance from the emitter, i.e. a close to constant power per surface area cone of radiant flux.
450mW is total output, observe the absence of an area unit. Unrealistically ignoring all losses you would need to spread it over 11.25cm^2 i.e. a ~3.7cm (sqrt(11.25/3.14) * 2)) or ~1.5inch diameter circle in an uniform way to get 40mW/cm^2
BTW given that the vF is 2V or so, the LED efficiency at 350mA is about 450/(350*2) = ~60%. We have come a long way from the days of incandescent light.
Unfortunately yes,they are as you say.I try to send these back,if possible,in a week or so,when my friend returns to NY. 
For what it`s worth a lot of the very Old LEDs (late 70`s) and Soviet era LEDs (the ones in the metal cans) are very deep red in color, when LEDs were first made it was hard to get them out the IR range and up into the visble part, and a good many of these early leds although visible as red go way down into the IR range too.
I wish I knew how to test them because I have several bags of these old leds and the soviet ones as well.
Hi Katherine Alicia,
First of all: Welcome to BLF! Nice to have you here. :-)
It's very interesting that these old LEDs, while being outdated and possibly low in efficiency, may get a chance for a "second life" after all. Do you know what type of LEDs these are? 5mm or another format? Can you provide a picture or even a spec sheet of them, please? Maybe UCL can chime in and tell us what they think about it. This thread has really catched my attention and I asked Sofirn to check with Epigap if there's a way to test the 6mW 5mm 670nm LED with a C01 host. Let's wait and see how this turns out.
Cheers,
Thomas
Thanks!
As far as I know the Russian leds are AL102 and AL310 they look about 4mm(ish), as for the old (western) LEDs, they`re in 5mm radial format but I`v no idea of any part number or anything?
When I get back to my parts bin I`ll take a pic and post them, I`d be happy to send some off to someone who can test them, just to make 100% sure they`re suitable.
they`re not like todays modern 5mm leds, they have a much deeper red and make most of todays 5mm LEDs look orange by comparison (620nm I assume?).
This is an interesting discussion. My brother studied electronics in the mid 80s, so I inherited a beat-up electronics tool box with a few vintage parts and some other scavenged thingamajigs over the years. Some old red led or two lying around (they're quite dim).
However, what is the aim of this? Unless it is possible to get a reliable source of whatever parts, and test them of course. I mean, I think a nice point would be to make a list of parts for a DIY build which anyone could build and use.
Forum members who can do spectral power distribution and other specialized led tests are maukka and djozz. Send them a private message if you will.
If someone already has a C01R could make some measurements please?
If possible measure the current draw in the 3 modes.
If possible measure the diameter of the beam against a flat wall or fixed surface from a distance of 6 cm (2-1/4 inches) and also at 10 cm (4 inches). From this we can calculate an approximate radiant power assuming all the energy is reflected and magnified by the lens and sent out the end. This might be a worst case to determine how much diffusion paper is required to cut the beam down to a safe level.
Or if someone has a radiometer can take measurements at these distances and modes, then we have it for comparison.
I hope you get the data you seek
It would be much simpler to just know the Lumen output of the light used for the study (including the diffuser), and then the lumen outputs of the C01R
from there it would be pretty easy to calculate the distance at which the two sources would have similar LUX on target, or to compare the output of the C01R with whatever diffusion is applied.
I find a couple of layers of scotch tape an effective diffuser…
I also find it relatively simple to move a light closer or farther away, to reduce or increase the visible brightness on target.
I feel I can pretty easily judge, what a dim level looks like. There is no need for excess brightness, the mitochondria dont need a specific dose… dim is plenty.