I got my light meter today but haven't had time to fully try it out. Is there some known standard for calibration purposes? Lux of a 100 watt incand ceiling light? How can I check my readings against some known standard?
When we give thanks to CPF. http://www.candlepowerforums.com/vb/showthread.php?94232-Light-Meter-Benchmark-Testing-%96-CPF-style
Yes, you can purchase some NIST ones. But is there a need? When the lights/emitter bins variations/drivers/electronics/batteries are not calibrated or standardised even.
The same thread contains a test for the cheapo meter many of us have: http://www.candlepowerforums.com/vb/showthread.php?94232-Light-Meter-Benchmark-Testing-%E2%80%93-CPF-style&p=2153108&viewfull=1#post2153108
It's actually more accurate on white and the spectra than most of the expensive meters except on blue.
The photopic standard is commonly the CIE standard observer.
Can you summarize what the test actually is. Maybe it's just because it's late but these 600 post threads are making my head spin :)
I see the page that you've linked to but I don't know how to use it. They mention A2 and white leds? How do I use that info?
Thanks.
A tad bit of background: a light meter measures not at a single point, but rather across a section of the EM spectrum. Consequently, a "complete" calibration for one means that its output is properly weighed across that section. It's essentially the information presented in a 2d graph, with frequency on X and relative magnitude on Y, specifically something like this: http://en.wikipedia.org/wiki/File:CIE_1931_XYZ_Color_Matching_Functions.svg. A measure of luminous light intensity (ie the kind that we can see, the kind we're concerned with) necessarily reflects one of these functions (eg photopic lux for the function just prior), and the output (the number shown on the meter) is then the sum of the output function after weighing against the input. A perfect light meter will weigh the entire spectral input against exactly that function, but in practice we only get an approximation. What we're investigating is how good that approximation is.
The point of the test @CPF is that testing across the entire spectrum is difficult/tedious, so instead the investigation is performed using various sample lights each with a different spectral input and ostensibly knowable output when weighed against that function above and summed (this was measured at some lab, ostensibly using calibrated instruments). Presumably the closer a random meter gets to this sum with its own, the "better" it is.
Each of the light sources are supposed to be something meaningful to measure. The A2 is presumably a SF A2 w/ some SF xenon lamp. The A1 I think is this: http://dmcleish.com/CPF/Aleph-1/index.html, with different emitters fitted. The specifics aren't really important, but the A2 is basically similar to a incandescent light bulb (black-body-like radiation curve), and A1 white is a typical white emitter the output of which you can see in CREE/Luxeon lit (blue spike w/ some shift to the redder parts). The other three are narrow band emitters at R, G, and B.
Anyway, the takeaway is that you look at the % diff to the calibrated number (the "LSI lab" one), with special attention to the spectral input you're interested in measuring, and it's good if it's close (on that specific meter). You can't compare your meter's numbers against theirs unless you get the exact kind of bulb/emitter they have (which is why they were shipping them around), though I supposed you can come close if you're very careful to buy something identical.
None of this info is that thread (I wonder if there was a precursor?), and it's kind of difficult to figure what's going on, in part just because it's so badly organized.
---
The practical solution to your problem above is that you basically need someone with a meter that's considered quite good, preferably consistently good across all sources, on that CPF test (ie it's very close to the LSI numbers), and use it to measure some sources that both you and the meter owner can acquire (like a known brand bulb / flashlight of reasonable QC). Then you can reasonably "calibrate" your number to his.
Is there anyone on here that has a calibrated meter? Dr Jones works in a lab and owns about every DX light there is :)
About the only light that I have that many people have is the Uniquefire HS-802. I also have the Trustfire Z1 (I think that's the model number...it's the CR123 angle light).
If Dr. Jones or anyone else with a calibrated meter is able to provide lux numbers at one meter for those lights it would be appreciated.
Another interesting thing for those of us with a few lights in common but who don't have calibrated meters would be to do the test anyway and post the results and see if they are all close.
I have other lights but I picked those two because they seem to be popular and I don't have any of the C8's that seem so popular. Drop-ins present problems anyway as they are all different.
I'll post my numbers for the HS-802 and for the Z1. I'll test 3 times to make sure the numbers are consistent but I'll only post one result for each. Batteries will be new or fully charged.
Uniquefire HS-802=14820 lux with freshly charged 18650 and tested right after turning on
Trustfire Z1 = 1205 lux with new battery and tested right after turning on
(without new or freshly charged batteries these tests won't be meaningful)
What actual distance did you measure the HS-802 from?
What current does it draw from battery?
1 meter...(39 inches)
1.01A
We can't compare then, as mine draws 1.4A.
My measurement, taken from 7.0m and converted to 1m, was 53k lux, but it's probably too high, also because I did it indoors and may have had some reflections from the walls. I need to redo it outdoors.
Also, I had the meter set to Max mode, moving the beam slowly, so as to capture the hottest part of the hotspot, in case there is any uneveness which the eye doesn't see.
Perhaps a longer measuring distance for the thrower would be more appropriate (even if your aim is not that of finding the light's true lux value), because of possible difference in focusing between various HS-802 samples. My converted 1m values showed a marked increase as the measuring distance increased, leveling out at about 5m.
What is the calculate I need to do to convert from 5 m back to 1 m? Thanks.
(are you using one 18650 or two CR123's?)
Multiply the reading by the m-distance squared. For 5m it would be x*25.
One 18650. I don't dare to put in two CR123's, it's already overdriven (from stock).
Do you recall what you got when you actually tried it at 1 meter?
Another good value to use which is derived from lux is throw distance. Eg a 60k light has a throw distance of 244.9 metres. Just square root it, what you are getting is 1 lux at x metres on the target. This is in accordance with the internationally accepted beam distance range values for searchlights.
How bright is 1 lux? Try it on your lux meter, not very bright.
I went outside tonight and tested the HS-802 at 3 meters and worked the number back to 1 meter and got 28,170 lux.
This seems to be more in line with some of the numbers I recall reading about in the past for this light.
I guess aspherics need more than 1 meter for the beam to fully converge.
I tried it with a diffused light (XP-G) at 1 meter it was 509 lux and at 3 meters converted back to 1 meter it was 333 lux.
I guess I need to experiment a little more. So far I haven't heard of anyone taking readings of any lights at 1 meter and 3 meters and getting close to the same number. That's a little troubling as to that particular technique. Maybe 1 meter isn't enough for the beam of any light to "mature".
That is why with big reflector HIDs, we try to do> 10 metres, 50 metres if you can. But if 12 metres is all you've got for the smaller HIDs, its probably ok. 5 metres will never make it. We are talking over a million lux @ 1m, the focusing is no joke....not like the "dim LEDs".
The centre hotspot of HIDs is pretty small. Really different beam profile than LEDs. Plus the 20k setting on your meter actually reads quite a bit different than the lower 200 or 2000 lux setting. In theory you may think that it is only 3 or 5%, but in reality it is really a very big difference.
The interesting thing is that aspherics seem to "improve" in their lux reading with distance (when converted back to 1 meter) but diffused light seems to show lower readings (when converted back to 1 meter).
You would think with a diffused light the reading at 1 meter and 3 meters would be the same or close. If it isn't there is some problem with the testing methodology or at least it should stabilize at some distance.
No, if you think about it, there is absolutely nothing wrong. What is the definition of "lux"? It is just the amount or intensity of light falling on an object.....at that particular free space or distance. Eg you put your sensor on the table, it reads 50 lux. Ok, that is good enough for reading. Just that.
It is us flashaholics 
who use it as a unit of measurement for determining "throw". You can do that, but as you've discovered, you need to be careful. We just convert it back to 1m just as a matter of convenience / basis of comparison. "Hey, i got 1205 lux @ 8.51m, and your short arc is 12921 lux @ 112.5 metres, that's cool!". What the!?!?
And for LEDs where the focus is not that critical (smaller reflectors, bigger emitters) it is pretty ok for a couple of metres measurement. Take a 6" HID host, pretty common size. That is a 150mm reflector. The arc length of an auto bulb is 3-4mm. What is your SST-90 SR90 thrower? 3mm by 3mm (9mm sq), 90mm reflector?
It is just a matter of focusing. Think of a magnifying glass, that would be an extreme example. You can burn wood at 0.5m as that is where the light is pin point. 1 billion lux. At 0.25cm that is 0.5 million lux. At 20 metres that is 20,000 lux.
Same for focusable laser pointers. Same for aspherics actually, if you fog up the beam with a head that is turned out/defocused you can see at perhaps 1m you can get say 500k lux for a Dereelight DBS aspheric with XR-E.
So moral of the story, just get out there to measure with longer distances, there is no problem at all.
With short arcs and precision rhodium reflectors they are talking in kilometres. With such a reflective reflector and focal point at infinity and arc length as small as 0.5mm, you need to be very precise to get the focus at infinity.
Anyway, the EmArc is supposed to be cutting edge now, brightest with the shortest arc and longest life. http://www.medicalbulbusa.com/ushio-smr-202-d1-200w-reflectorized-emarc-lamp.html
Maybe I've missed your point but how do you explain a diffused light giving different "converted" numbers at 1 meter and 3 meters. Surely it has reached it's focus at 1 meter (it's diffused) so the 1 meter numbers and the 3 meter number (converted back to 1 meter) should be the same shouldn't they?