Measuring Lumen Output — How to For Beginners?

Okay guys, be gentle; I'm not a flashlight tech-head by any measure LOL. (Also hoping that this is the right forum for this topic; mod please move if not.)

Is this little lux meter a practical and cheap way to check lumen outputs of my meagre collection of flashlights:

I'm familiar with the basic ideas of lumen levels, throw, spread, basic LED technology, reflector/lens types etc but not really with measuring OTF lumens. Which I'm guessing is the only meaningful way to compare flashlights — ignoring for the moment their intended purpose... or manufacturer's advertising blurb.

The sort of questions I have are like do I measure the lux at 1m at the very edge of the spillbeam and at the centre of the hot spot? Or Neither? How do I measure it at a specific throw distance, say 50m? Or do I do all the measurements at 1m and then simply interpolate the result? Or am I grossly oversimplifying the use of this lux meter altogether.

Also, if there's a convenient site or page here that answers these sorts of newbie questions, then I'd appreciate it if someone here could suggest one. It'd also save you guys wasting your time explaining this stuff LOL.

—Thanks, Geoff.

The best way would be to make your own integrating sphere. Aside from that, another fairly general way to compare lights is by doing a ceiling bounce test. What you do is go to a room that has no ambient light and a white ceiling, and preferably white walls (bathrooms are usually good for this). If you already have a luxmeter, then place it on the floor with the detector pointing up. Then place the torch on the floor and turn it on while it's pointing up at the ceiling. Record the number in lux. If you don't have a luxmeter you can still perform this test by using two torches back to back...You'll be eyeballing it, but it's a good quick way to tell which torch is putting out more light, regardless of beam style. Also, if you have a good camera you can take photos of the the ceiling bounce tests, then compare them on your computer screen. Check out Foy's excellent post where he compares a few torches this way.

Hope this helps,


I too an curious about this and want to learn. Thanks for asking about this.

Thank you muchly for that link Match... and of course your great "recipe" for actually making the sphere. It's illuminated me no end... ouch.

One more question. These are the specs of that lux meter I've linked to.

Measures from 0 - 50,000 Lux/Fc in four ranges

0.1 Lux/Fc on 0 - 200 range, 1.0Lux/Fc on 0 - 2,000 range

Are these adequate?

—Thanks again, Geoff.

I find that using the room integrating sphere isn't reliable w/ the meter pointing up. Seems to vary too much based on the optics, just like you don't want the meter directly opposite of the hotspot in the sphere. So I point it at a side wall.

You can make a simple IS with a milk carton.

The cheapo luxmeter(s) on DX work fine.

hehe...pretty punny! Yup, that range should be just fine on that meter.

Ah, I point the lux-meter directly up at the ceiling, too. Pointing it to the side wall weights parts of the beam near that wall much more than others...

Also I don't really trust light boxes....

ausGeoff: A luxmeter without an integrating sphere still lets you measure the spot intensity and the throw. I gathered some info here.

On some of my lower-output lights, I can stare straight into the reflector on "high" for a full second, and my vision returns in about 3 minutes.

With XM-L's driven over 2.5A, it takes about 7.5 minutes for my vision in that eye to return, but with the 1A-driven vShark, it comes back in 6 minutes. Only use one eye, you need the other for stopwatch or you won't know how long you have been blinded.

Anyway, that method seemed to work ok for me until I got my 200mW green laser, since which time I've been telling my kids I'm practicing for a pirate party with the patch.

LOL, ya know I'm just kidding, don't try that at home!

The integrated eyeball test.

"Only use one eye, you need the other for stopwatch or you won't know how long you have been blinded.

Anyway, that method seemed to work ok for me until I got my 200mW green laser, since which time I've been telling my kids I'm practicing for a pirate party with the patch."


LMAO! Brilliant, trooplewis! Kind of like how the voltage of an electric fence is directly proportional to the size of the wet spot made on the front of trousers after being touched :)



That's a CEM DT-1300/1301 (think on the Aussie page they blanked out the logo LOL!). Do note that they consistently under-read by some 10-15% in lux for LEDs, not that they purposely wanna do that, just that its calibrated to a 2856K incan source. CW LED has a drop in yellows in the spectral output. Well documented in CPF.

Look at the DT-1300, its very nearly 100% accurate for incan.

BTW when you are mentioning lux, you need to mention distance. Or you can mention Candlepower. Eg 10,000 lux at 10m = 10000 x 10^2 = 1,000,000 1 million candlepower. (a lot of those searchlight have very inflated CP figures). This gives you a very good idea on throw, though not on corona/hotspot size and characteristics.

HIDs apparently do not have so much of an issue, even for 5000K bulbs.

Thanks DrJones for the link to your page...

Lots of useful reading for me there. (As well as the "throw" page.)

—Cheers, Geoff.

I just make sure the geometry is such that the meter doesn't catch any spot directly illuminated by the optics, only reflected light. When I tried the ceiling method, try on different optics would vary the readings pretty drastically (esp w/ diffusers and such).

Cool. Kinda like flashlightes 101. Thanks for starting this topic!

Ever seen a dog that watered an electric fence? It is utterly hilarious seeing a dog trying to walk around with crossed legs. But I'm glad it wasn't me that got thousands of volts through my prostate.