How red is red?

I've been thinking about some of the ins and outs of scotopic vision and I'm trying to figure out what wavelength would be optimal for an LED to BOTH preserve dark adapted vision AND still be useful. The eye seems to be pretty insensitive to real red (above 655nm).

Does anyone know the wavelength(s) of the LEDs that the various light makers use in their lights that are meant to preserve night-vision? The only one I know the wavelength of is the NightCutter Chartlight, which is 626nm.

Anyone know the wavelength of the red LEDs in the Rigel Systems Skylight or that of the LEDs in any other red flashlights? Anyone know what wavelength the US Navy calls red for preservation of night vision?

I've also read about using a very dim white light for reading charts. While this would solve the obvious color recognition problems of the red light, can night vision truly be preserved even if one is using a very dim white light?

Please excuse these dopey questions, but I've been looking over photopic/scotopic sensitivity graphs and there are just some things that don't make sense to me.

I think the theory is that your rod type retinal cells (which can't detect color) are what provide you with your night vision. In low light, they can increase their sensitivity up higher than the cones (color detecting cells). You have 3 kinds of cones which roughly detect red, green and blue which have different spectral responses. Apparently the 4 types of cells each adjust their sensitivity more or less independently to the light spectra that they see.

It turns out that the spectral resonse of the red detecting cones extends out lower (longer wavelengths) than the spectral resonse of the rods, so that red light below a certain frequency can be detected by your red cones without causing your rods to reduce their night vision sensitivity as the rods - by and large - don't see that red light.

How red the light has to be depends on the spectral resonse of red cone cells vs. the rods. Here is a reference

http://www.yorku.ca/eye/specsens.htm

which shows that a light with most of its spectrum longer than 570 nanometers or so (red cone detection peak is at 564 nm) should be usable by the red cones without affecting the night vision sensitivity of the rods.

This is my vague recollection on the subject. Anyone finds this not right, please correct me.

So many people have their own theories about works best for night vision. Some say red, some green, some orange, etc.

I believe in the red argument simply from experimenting from turning on my Arc RGB in complete darkness. When this is done, I can't for the life of me stare and the green and blue shining on the wall without squinting from them being "too bright" but for some reason, I can just stare and stare at the red and don't squint at all. I personally believe that higher intensity reds in lower wavelengths, below 635 nm, and if possible even 650 nm, are better then less intense lights at higher wavelengths much above 635 nm. This is what always seems to work for me, but I'd say you cutoff line is around that area. 626 nm should be acceptable as long as there isn't a lot of spill into 610's.

Here's a snippet from another post in a different thread..


Night vision is the result of the buildup of rhodopsin (commonly called "visual purple"). This takes 30 to 60 minutes in complete darkness. DIM red light has no effect on rhodopsin. Any other color at an intensity where you can see the color (not just identify it as "light") destroys it and it has to build up again. Hence why dim red light is best for preserving night vision. Green light is used by the military for night vision equipment which is less sensitive to green light. However green light will break down rhodopsin-based night vision instantly at an intensity where the color of the light is recognizable.

I thought the subject important enough, I spent a lot of time researching and came up with my web page Night Vision - The Red Myth.
There are also threads in the Night Vision section here along with the ones on assisted night vision.

nerd,

I'm going to have to spend some time pouring over that site -- thanks for the link.

I'm a bit confused as to why a wavelength as short as 570nm would work for this. Being on the green/yellow border, 570nm would seem to be a FAR more usable color than red or even any orange in terms of sensitivity.

Wow, two more posts while I was writing my first response.

It is comforting to note that there is some disagreement on what is actually going on with night vision. The more graphs I study, the more diverse the opinions become. At least it's not just me.


gwbaltzell,

Thanks for the link. I'm going to go over that one, too.

Hmm.. more to it than I thought. Nice collection of data.

570 nm doesn't work if the goal is to completely not involve the rods. Must be 650 nm or longer and then only very low level. The truth is in most situtations you do want to use the rods for what they were intended - night vision. However, people will argue this all night long.

[ QUOTE ]
gwbaltzell said:
I thought the subject important enough, I spent a lot of time researching and came up with my web page Night Vision - The Red Myth

[/ QUOTE ]

The correct color - as identified in the opening post - to preserve true Scotopic night vision is Red.

This is backed by the flight surgeon of the USAF -

Red, Green, Blue-Green, or White?

from -
USAF Flight Surgeon's Guide
Chapter 8
AEROSPACE OPHTHALMOLOGY
Thomas J. Tredici, M.D.

http://wwwsam.brooks.af.mil/af/files/fsguide/HTML/Chapter_08.html

QUOTE
Cockpit Illumination: The use of red light (wavelength greater than 650 nanometers) for illumination of the cockpit is desirable, because it, like red goggles, does not affect dark adaptation. Red cockpit lighting has been traditional since World War II. The intent was to maintain the greatest rod sensitivity possible, while still providing some illumination for central foveal vision. However, red cockpit lighting did create some near vision problems for the pre-presbyopic and presbyopic aviators. With the increased use of electronic and electro-optical devices for navigation, target detection, and night vision, the importance of the pilot's visual efficiency within the cockpit has increased and new problems have been created. Low intensity, white cockpit lighting is presently used to solve those problems. It affords a more natural visual environment within the aircraft, without degrading the color of objects. Blue-green cockpit lighting is used in aircraft in which night-vision devices are used because, unlike the human eye, these devices are not sensitive to light at that end of the visual spectrum. In addition, blue-green light is the easiest for accommodative focus and is seen by the rods more readily than any other color. It is not seen as blue-green, however, but only as light. However, the enemy can easily see a blue-green light, under scotopic conditions, in any position of his peripheral field, whereas a low intensity red light would be invisible unless viewed directly.
UNQUOTE


We also had a long discussion about this in this thread - read all the different opinions to your heart's content /ubbthreads/images/graemlins/smile.gif

Preserving Night Vision - Colors?

UnknownVT,

I'm going to check out both links.

Thanks

Here's another useful Pinned Sticky Thread in the Night Vision section of CPF -
Human Night Vision Preservation

But wait, isn't the point of CPF to not NEED nightvision? /ubbthreads/images/graemlins/smile.gif

Only thing I can really say in response to this, is that when I was converting the instrument panel illumination in my jeep to LED's. I tried a few colours. (White, green, yellow green, yellow, amber, and red). The red (which I am still using) offered by FAR the best balance. Being bright enough to see, without introducing any glare or interfering with my view of the road, even with only dipped headlights on. Amber was next, the glare problem just got progressively worse as the wavelength decreased.

[ QUOTE ]
nerdgineer said:
But wait, isn't the point of CPF to not NEED nightvision? /ubbthreads/images/graemlins/smile.gif

[/ QUOTE ]
Alas...in a perfect world...

Wow, thanks for all the input. After a few responses I realized that my subject line was really lame, but you all found me just the same.

I followed all of the links, threads and secondary links and I was sort of blown away by all of the directions it went. This is obviously an example of a question that defies a simple answer.

Thanks again

[ QUOTE ]
Zelandeth said:
Only thing I can really say in response to this, is that when I was converting the instrument panel illumination in my jeep to LED's. I tried a few colours. (White, green, yellow green, yellow, amber, and red). The red (which I am still using) offered by FAR the best balance. Being bright enough to see, without introducing any glare or interfering with my view of the road, even with only dipped headlights on. Amber was next, the glare problem just got progressively worse as the wavelength decreased.

[/ QUOTE ]

I didn't convert my instrument panel, but I use a colored LED in my car while driving at night - especially when I make it to the mountains. I've been doing this for about 4 years. Red and amber seem to me to be the least obtrusive. Dim green is OK. Cyan bothers my eyes.

What I still don't know is whether that has to do with NV, or whether it just has to do with your eyes being bothered by shorter wavelength light.

Scotopic vision has very low resolution. You can't really use it for making out fine details, like for reading charts. Its purpose is to let you avoid bumping into large objects when walking around at night, to see where other people near you are, etc.

You can get a good demo of scotopic vision if you sleep in a pitch dark room and you have a GITD wristwatch (the kind with luminous paint on the markers, not tritium and not an electric backlight). If you wake up in the middle of the night when the GITD charge has mostly faded, you'll still be able to make out the hands and markers, but they will look white (even though they're greenish as can be seen when they're brighter) be quite blurry. That's scotopic vision.

To get an impressive demo of how sensitive it is. try covering the bezel of your Arc AAA with your finger, then turn the light on so all the light is blocked. Then carefully uncover just the tiniest sliver of the spill beam. That's like one hundredth of the full output of the AAA, and yet you'll find that you can see all over the room with it.

I think there are two different methods one can take to "preserve night vision."

1) Use the wavelength that is least disruptive to dark-adapted eyes. This seems to be somewhere in the red range.

2) Use the wavelength that the eyes are most sensitive to, so that the raw number of photons used can be minimized. In this case, green/cyan seems to be the winner. I guess the theory here is that the eyes are so receptive to this wavelength that you can use a very small amount of light such that dark-adaptation is preserved.

With regard to case 2), it may also make sense to use as dim a white as you can get away with.

I used to do long throw comparisons of white vs. cyan, and even though cyan appeared brighter, the improved color contrast of white allowed most objects to be identified more clearly. In other words, it's not just brightness contrast, but also color contrast that allows a target to be identified.

So if you can make out objects better with dim white than with dim cyan, you should be able to dim down the white more than the cyan, and that would support white being a better candidate for case 2) than cyan.

I can't say I've tried it for NV yet, though. However, I no longer sneer at the dim white crowd like I used to. /ubbthreads/images/graemlins/smile.gif

[ QUOTE ]
Lux Luthor said:
With regard to case 2), it may also make sense to use as dim a white as you can get away with.

[/ QUOTE ]

Let me throw my other 2 cents into this - note I am no expert.

I am an advocate for DIM lights - or rather the right level of light suited for the purpose.

I do understand the use of as dim a light as one can manage - and one would be amazed how DIM that is when one is actually using true Scotopic night vision - that is Rods only vision.

Switching on almost any light will relatively speaking "dazzle".

Look at that last sentence again - this means probably almost ANY non-custom/modified light will not be dim enough.

So using any other color than Red is going to affect the Scotopic night vision by flooding the Rods.

Now unless one is an astronomer in a sensitive obseratory (try turning on any other colored light than red there during their blackout, and see what happens to you.... /ubbthreads/images/graemlins/grin.gif) -
we really probably are NOT using anything like true Scotopic night vision - what we have is some "dark adaption" - where it is true that any DIM light will help preserve that "dark adaption" - hence the very valid point (2) refered to above for NON-Scotopic "dark adaption" - Cyan, white or any color that you like as long as it's dim....... I personally like yellow/amber because I can see well under that color for definition and contrast, albeit with the loss of color rendition - and it disturbs my (NON-Scotopic) dark adaption little.

I also have a dim white coin cell light (20 for $20 from CountyComm running on 3V - single 2032 or 2016 with dummy battery) that I found useful - but because these are dim - I think I see better with the yellow (other than color rendition) ......

See:
joys of a DIM flashlight