by Gordon Bond
Originally appeared in TPO Volume 10, Issue 2.
©1999 Typographica Publishing.
May not be reproduced for commercial purposes.
If you ask the average amateur astronomer what color light they use to see by
when observing, the answer will almost invariably be "red". But ask them why red is
best and they will likely be unable to give a really satisfactory answer. Why is it
that red, above all other colors, has become the standard "astronomer’s flashlight"
I’ve been interested in this subject ever since hearing that a few observers
were challenging this conventional wisdom. I’ve already published two other article
on the matter in two consecutive issues of TPO ("Red or Green", Volume 7, Issue 3,
and "The White Light Heresies", Volume 7, Issue 4). So, why publish yet another
article? The flashlight is a universal and fundamental tool of the visual observer.
How well it performs is critical to how well we perform as observers. Also, there
appears to be plenty of conflicting evidence showing that this is by no means as
simple an issue as might at first be supposed.
This whole thing started back in 1995 while attending the third annual United
Astronomy Clubs of New Jersey Symposium. Several members were discussing an email
thread that suggested the possibility of green light being better. CCD imagers had
been using green LED flashlights since their CCD chips were less sensitive to green
than to red. Some of these users felt the green light was better for preserving
dark adaption of the eye as well.
It was Dr. D.T. Durig, Director of the Cordell-Lorenz Observatory at the
University of the South, Sewanee, Tennessee, who started the thread. "The problem
with color and the human eye response in astronomy" he wrote, "is that the color
receptors in the eye are cones and when you become dark adapted you do not use the
cones but instead the rods. The rods allow you to detect dimmer variations in the
light intensity, but do not have color sensitivity."
The human eye is more sensitive to the green wavelengths of light than the red
or blue. The logic became that if the eye is less sensitive to red, more red light
would be needed to see by. With more red light, the more obvious the red coloration
would become. If, by definition, dark adaption involves using the color-insensitive
rods, the fact color can be perceived means that the cones are involved and dark
adaption is compromised.
Since the eye is more sensitive to green, the line of reasoning went, you could
use dimmer light to see comfortably and the green color would be less obvious. My own
informal experiments with a green LED penlight showed that by pulling the light far
enough from a sheet of white paper, it appeared to cast an almost grayish glow -
enough to read by, but not enough to show the green color.
This idea was supported by amateur CCD pioneer Richard Berry, "The idea is to use
very dim green LEDs. I have experimented with these and found that very dim green does
less damage to deep dark adaption than the standard red LEDs, which are hard to see
with and almost always much too bright. The level is right when the light looks grayish
As it turned out, however, the matter is not so simple. In fact, the basic premise
may be flawed. I brought up this subject this summer as part of the deep sky egroup,
email@example.com and began yet another thread. Lowell Observatory’s Brian Skiff
weighed into the fray citing several paper titles that had appeared in the Journal of
the Optical Society of America (JOSA). "The first lesson I got from looking at
the...papers we have" he wrote, "is that when one is dark-adapted, your spectral
sensitivity is not purely scotopic, as the standard lore would have it. It is instead a
mix of rod/cone sensivities even at quite low light levels. Mixed in with this is the
eye’s tendency to alternate between ‘centrally fixated’ (i.e. direct fovial) vision and
good ol’ averted vision, which people tend to use naturally in the dark. The fact that
there are still papers getting published on the subject (the last reference is to a 1998
publication) means that the whole business is rather more complicated than at least my
simple view (now amended) would have it."
In short, Skiff is saying that the evidence shows that dark adaption isn’t a
matter of using either cones or rods exclusively, but a more complex combination of
the two, though the rods are likely the more active.
Skiff also claims that "there is ample laboratory evidence to show that red
light allows you to become dark adapted after exposure (even at very low levels) about
4 times faster than with green or white light."
Jeff Medkeff, the Acting Assistant Coordinator for the Association of Lunar and
Planetary Observers, Solar Section, joined in the discussion in support of Skiff’s
claims, citing the specific JOSA papers (listed at the end of this article). "Brian
is totally and completely correct" Medkeff asserts strongly, "to say that you get the
dark adaption back quickest using red light at the intensity necessary for reading,
then any other color."
Even if there is evidence that red light overall allows faster recovery to dark
adaption, there is another consideration. Obviously, we use our lights to do a specific
task - read. Whatever color, we must be able to read text in catalogs or symbols on
star charts. This may be an area where red has some difficulties. Surprisingly, some
observers advocate the use of white light.
Walk into a star party with a white light flashlight, albeit dim and well-shielded,
and you’ll be taking your life into your hands. Nevertheless, there has been a school
of thought that a very dim white light will serve better than colored light. The human
eye evolved primarily for use in the relatively white light of the sun. Therefore, it
is most sensitive to white light. Part of the thinking behind using green light was
that because the human eye is more sensitive to green, you’ll need less of it to
comfortably illuminate something. The white light supporters take this one step further.
Because the human eye evolved to work best in white light, the sensitivity will be
highest when using it and even less of an illumination level will be needed.
A LIGHT TO READ BY
Doug Kniffen’s informal experiments and personal experience have made him a
supporter of white light. His argument centers on the issue of contrast. "Two
qualifications are required for the eye to see: sufficient illumination and, very
importantly, sufficient contrast." he says, "The minimum necessary illumination
level will depend more on contrast...Both the color of the light source and the color
of an object viewed by reflected light will affect the level of visible contrast
needed for adequate perception to occur."
Kniffen maintains that white light creates the best contrast and therefore
readability. "Since an integrated light source is composed of all colors" he writes,
"white light is the best color to use when trying to see with the minimum necessary
intensity of illumination."
He’s not alone. While researching light pollution data in the Department of
Commerce’s National Technical Information Service, Robert Bunge came across an
interesting paper. The 1986 paper, published by S.M. Luria and D.A. Kobus of the
Naval Submarine Medical Research Laboratory in Groton, Connecticut, seems to support
the idea that red light isn’t necessarily the best to read by. According to the
abstract, "Red light has been used for many years to illuminate various submarine
compartments at night, because dark adaption is then achieved more quickly when the
ambient light is turned off. This one advantage of red light is offset, however, by
several disadvantages. The red light is fatiguing, it makes it impossible to read
color-coded charts, and indeed, it makes it more difficult to read anything...A long
series of studies has now shown that the problems associated with red light would be
alleviated, if not eliminated, by substituting white light of generally comparable
The needs of the submariner are, of course, somewhat different than those of
the amateur astronomer. Jeff Medkeff countered the argument, "Endless discussion of
what is put in airplane cockpits, submarines, or nuclear reactor control rooms is
totally irrelevant to the question of visual astronomy...if you want to look at your
telescope at night, go with the green or the ‘passband white’ that is being put on
submarines. If you want to look through your telescope at something close to the
limits of vision, you better optimize the lighting for that task."
Medkeff’s argument is that the primary requirement of the pilot, submariner or
others mentioned is to be able to see their instruments. They don’t use averted
vision and are not trying to see threshold objects. I would counter, however, that
amateur astronomers need to do both. We need to be able to adequately see, say, a
finder chart and compare the view with what’s in the eyepiece, or we may want to
make a drawing or take notes while looking at the object.
The submarine report does bring up another potentially important issue. Reading by
red light may increase eye fatigue.
Nils Olof Carlin conducted some interesting experiments using red, green and yellow
LEDs. He described it as follows, "I had previously made a box with batteries and
LEDs of different colors, red, yellow and green, and switches to choose one and
turn it off and on. The LED plastic ‘lenses’ were all ground flat to ensure even
"I took a text with about 4mm font size and a ruler, and sat down in the darkest
room of the house, the chart on my lap. I held the ruler, one end on the chart
and a marking beside my eyes, fixing the reading distance at about 650mm. Then I
held the LED box to illuminate the text and moved it along the ruler until I found
the distance where the text was barely readable. I repeated this a few times for
each color, but the distance turned out to be well reproducible (to avoid
differences in focusing of different colors, I ‘focused’ with the progressive
glasses for each reading)."
"For the next phase, I put a piece of black flocking paper on the floor, and put
2 pieces of white paper, about 8 and 16mm. After some extra dark adaption (only
needed in this phase), I lifted the LED to the height where I barely could see the
larger [piece of white paper] on the floor by averted vision (reading distance here
was about 1500mm, meaning angular sizes of 20-40 arcminutes) - again, repeating
this a few times for each color."
"Now I had ratios of the distances used in the first and second phases, and the
inverse of the square of this ratio is the ratio of illumination needed for
chart-reading (direct vision) and spot visibility (averted vision) respectively."
"The ratio of illumination I needed for test text to test object (averted vision)
was (well-rounded), 300 for green, 200 for yellow and only 20 for the winner, red."
While Carlin’s results lend support to the use of red light over green, they don’t
address white light. But he does not claim his work to be the last word and hopes
"someone else would try it and offer their independent results."
Several experiments seem to show that for astronomical purposes, red light is the
best choice. Nevertheless, the practical "in the field" experiences of some
observers seem to show that green and while light may also have merit. Why such
divergent opinions? It is possible that there are factors involved in actual field
use that are not being reproduced or accounted for in controlled experiments.
Perhaps the greatest variable is the human eye itself. The age and health of the
individual plays a roll. The width the iris opens under dark conditions becomes
less with age, for example. Certainly nearsightedness or farsightedness must be
a factor. Individuals under stress may have poor blood-flow into the eyes. And
what about temperature? How might extreme cold effect the circulation of blood
to the eye? Brian Skiff commented that there is a point "where things like
breathing and just being calm enter into threshold sensitivity as much as dark
Don Pensack made an interesting observation regarding the human factor of age.
"I use a red LED light with the brightness turned down to record my notes,"
he wrote, "but I’ve noticed I have to use reading glasses of higher power than
I normally use to see in such dim conditions."
Jeff Medkeff concurs. "Me too - I noticed this when I moved from Ohio to Arizona
and started to look at dim stuff again. I was concerned at the time because when
I was a teenager in Ohio, I never needed reading glasses with my red light. Now
that I am - well, not a teenager - I find that the difficulty of focusing red
light is enough that I use a mild diopter reading glass from the drugstore to see
Another factor in the equation is that all LEDs are not created equal. Doug
Kniffen studied the spectral emissions of many different colored LEDs and found
"The vast majority of LEDs are not monochromatic! Some of the red ones will emit
more blue light than some of the green ones do. Their apparent red color is
determined by a combination of peak emission wavelengths and overall spectral
emissions. A yellow LED will frequently have more red and blue in its spectrum
than a green LED."
KEEP IT DIM
There seems to be little real consensus on this matter, but there is one point on
which all seem to agree. No matter what color light you prefer, keeping the
intensity level low is the key.
Richard Berry remarked in the earlier discussions that the green light he
advocated should be kept at such a level
that it appears grayish. Red advocate Brian Skiff wrote, "I think it was Roger
Clark who gave as a rule-of-thumb that if the light is bright enough that it makes
the paper you are looking at red, then it is too bright. I have a crummy red LED
light that I can turn down so that maps, etc. appear a funny grey-brown."
Another point that most people in this discussion seem to agree on is that the
majority of the red astronomers flashlights encountered on the market or being
used in the field are much too bright. Jeff Medkeff commented that "at
the Texas Star Party earlier this year, the only person I saw (outside of my own
party) using a properly-dim red flashlight was Jay McNeil - and that is out of
how many hundreds of serious, experienced observers?"
All this discussion may seem rather academic to some. After all, haven’t we
gotten along just fine for decades with the red flashlight? But for serious
visual observers, this can be a critical issue. We sometimes take great pains to
try and stack the odds in our favor when hunting for some very faint object or
subtle detail. We try to optimize the performance of our telescopes to squeeze
out every photon possible, we travel great distances for darker skies and wait
for the ideal weather conditions. Preserving our dark adaption is just another
aspect of this process and the gain or loss, no matter how small, can sometimes
make all the difference between seeing something and not.
What is needed is to design an experiment which will take into account as many
of the variables as possible to explore if red light is really as good as
it seems for all observers. TPO is willing to take up this challenge. I am asking
my readers to give their comments or suggestions as to how such a definitive
experiment might be conducted. E-mail me at firstname.lastname@example.org or write to me at
the address listed in the [contacts section].