Take these results with a whole bag of salt because these numbers are guestimates after extrapolating a factor from lights that I have of known OTF lumen numbers and then multiplying that with the lux figure that I got. Note that, my lux meter is my trusty Nokia N900. The light meter on that phone seems to be more sensitive than my HTC One S. All lights were tested at their respective high/turbo settings unless otherwise stated. Also, all the lights were using either NCR18650A or NCR18650B for 18650 based lights. Duraloops were used for AA based lights. The cells were fully charged before testing.
Light
Lux
Calculated Lumens
D40A-NW (medium)
3
234
D40A-NW (high)
7
546
XinTD X3
7
546
M3C4 (XM-L, original)
8
624
OEM Big Head (modded)
8
624
Courui/Big Head Torch (XP-G2)
8
624
Courui/Big Head Torch (XM-L2*)
10
780
D40A-NW (turbo)
12
936
DRY (original driver)
42
3276
SC6330 (Vinh modded, NW)
45
3510
BTU Shocker (Vinh modded, NCR18650PF)
61
4758
*Might need to test this one again. I thought I had dried out the light after yesterday’s less than fortunate run-time test, but there was still water in the light. Will probably try again in a couple of days. I’ll update the OP if there are any changes.
To my eye, these are about right. In that, if one scored a “7” and the other, an “8”, it was visibly brighter than the other light. If they had the same score, again, to my eye, they were equally bright. As for the lumens, well, as I mentioned at the start, that’s an uneducated guess.
Ahhhh, smartphone apps as light meters… they work quite well. I’ve been metering my lights for while now with a DSLR and have tested a smartphone app with good results (although the DSLR meters are more sensitive and accurate). Anyways, here are a few ideas that you might find helpful.
- bounce the light horizontally in a narrow hallway (walls ~3 ft part) - bounce off one wall, read off the other. Closing the bounce/light meter distance will great improve brightness for more accurate/consistent readings.
- your whole number lux readings with 78 lm increments are kinda wide, esp if you want to measure lower lumens, the horizontal bounce will help some, but consider using a photography light meter app (with EV, shutter speed, aperture, ISO) which will have finer increments.
- two ways to use photog apps to read lumens, first and simplest is to calibrate to and solve for shutter speed (eg, calibrate meter for 100 lms @ 1/100 of a second). Second, and with much finer increments, is to convert EV (exposure value) to lumens with a formula (happy to provide, if you get to this point).
- finally, you can make a mini lightbox, borrowing the 3 piece PVC plumbing elbow a few folks around here are using. You can accurately measure down to sub-lumens with all the light contained.
Thanks for the tip. The setup was intended for my brighter lights. I had no intention to use any of my sub-200 lumen lights as part of the test.
On a secondary note, I did do some tweaking to the XM-L2. I stacked on another R120 resistor on top of the existing R100 along with changing the wires in the driver to 18AWG and it was brighter. The initial start-up was an “11”. However, a few seconds later, it settled down to a “14”. I would suggest that this number would indicate something in the 1100 lumen region and this was verified by confirming by eye, that the XM-L2 was brighter than the D40A-NW on turbo.
