lux, candela - proper terminology

This is a plea for using the proper units (and physical quantities) to describe the performance of a flashlight.

Often one can read about "lux" or "lux@1m" to describe 'spot brightness' or throw.

lux is the unit of illuminance E, that is the luminous flux F (lumens) per area A falling onto a surface. It's more a unit for a whole lighting setup onto a target than for a light source itself.

With a single light source like a flashlight, the illuminance E drops with increasing distance d, described by the inverse square law: E is proportional to 1/d² (actually that's an approximation which gets more accurate with increasing distance). If one now multiplies that illuminance E by the squared distance, E*d² gives a constant value, which is independent of the distance and only depends on the flashlight itself. This is another physical quantity, the luminous intensity I, or, measured at the spot, the spot luminous intensity I, or peak luminous intensity I. It's unit is candela (cd; candlepower (cp) is an obsolete unit equivalent to candela): I = E * d²

Of course, one lux in d=1m distance gives one candela, which is why "10000 lux@1m" is somewhat the same as 10000 cd (or better 10 kcd), however the latter is the proper unit. ANSI/NEMA FL1 uses it, too.

Most of you already know that measuring the spot intensity at only 1m distance is not a good idea, as the beam often has not settled to it's 'final beam pattern' in that short distance. Even ANSI/NEMA FL1 demands a minimum distance of 2m; for throwers that is still much too near (I suggest a distance where the spot size is at least 10 times as big as the reflector/lens diameter). Doing so and calculating that back to "lux@1m" sometimes leads to strange units like "31736 lux @1m (calculated)" - just write 31.7 kcd instead.

I just would like to suggest

  • not using lux when you mean candela
  • using the "k" for big numbers
  • not denoting more than three significant digits; typical lux-meters have an accuracy of 5%, a number like 31736 seems to imply a dubious precision of less than 0.1%

While being at it, throw (in meters) is usually defined as the distance at which the illuminance on a (perpendicular) target falls below a certain level; if you follow ANSI/NEMA, that level is 0.25 lx. Throw can be calculated from the spot intensity I using the above formula rearranged to
d = sqrt ( I / E0 )
with E0 = 0.25 lx for NEMA throw. (Some consider that value of 0.25 lx quite low, resulting in relatively high throw values; practical throw values might be lower, depending on the situation.) (more info on throw)

You might find this a bit nitpicking; but I think proper terminology is a good thing, it was developed for a reason.


Hello Dr Jones,

You seem to be one of the many people here with some serious skills.

I am kinda new around here and am learnng a lot and I would like to measure what the lights I have are really doing in the photon department. I am an ANALytical person so if I try to take measurements I would like them to be reasonably accurate and meaningful. I can do current and voltage measurements at the same time and I have a lux meter of reasonable quality so if you were do define a test procedure what would it be? I have had some thoughts but as I read more here I realize how much that I do not understand. I am not looking for laboratory grade, just something that is reasonably simple and repeatable.

Thanks for any information you can provide.


Thank you DrJones for this attempt at putting us straight.

Some time ago "bro 2100" motioned toward this already, without going into so much detail. So I tried to get a clearer picture of that all but got confused by another unit of similar name instead (I won't mention it so as not to confuse others too, but it was the other obsolete one with the factor 10.764 - what the heck?) and I finally ditched the whole thing and returned to the popular lux@1m. These photometric quantities are quite a bitch to comprehend, I must say that.

Anyway, I feel ready to adopt the candela now.

@Larry, that info was 'hidden' in the link I included... :)

Perform the measurements in a dark room (check background level) and avoid reflections from table/floor/walls (including diffuse reflections). Using a cardboard aperture is advised.

Choose a sufficiently long distance; I usually use 3.162m (guess why), but more for very throwy lights. I once used 11m for a 100kcd thrower and 30m for a 400kcd thrower. When in doubt, measure at two different distances and check for consistency. If inconsistent, try longer distances.

That said,

  1. use a lux-meter to measure the maximum illuminance E [lux] in distance d [m] from the flashlight (but ignore the first 30 seconds, according to NEMA),
  2. use I=E*d² to calculate the spot intensityI [cd].
  3. Then use d=sqrt(I/E0) (with E0=0.25lx) to obtain (NEMA-)throw [m].


Thing is, if you are using longer distance, it'd already take > 30 seconds when walking to that spot, so no worries on whether the light has already stabilised or not. LOL! :D

This is not totally off topic...I just noticed a product spec on DX stating the lux as 1300 lux @ 1m.

Ok, according to this thread it should simply be 1.3 kcd but I've never seen them provide lux in their spec for any product before. Is this the dawning of a new era..maybe the rising of Aquarius!

I'm seeing DX advertising banners on websites all over the internet lately as well. It's a whole new world :)

DrJones, thank you very much for this timely write-up! When my cheap eBay light meter gets in, I will be using it to check the illumination on some of my lights. I was online last night reviewing information on light output for throwers (including your CAT100 lamp Smile) and was wishing you or someone else with your level of understanding had a good document with advice on measuring throw. I'm not sure how I missed this over the weekend.

Any new advances in thrower tech since the CAT100?

Actually yes: the FAT NEEDLE 400... (I had posted that at CPF)

However I won't be able to surpass Walterk's Huygens Ultimate, since I don't have such a big lens...