Calibrating Light Meter-standard?

LED Flashlights – General Info
21

Try 4 metres and 8 metres. Try also to measure as low as possible, either in the 200 or 2000 lux range. You cannot be near to any reflective object as well for both light as well as sensor location (eg wall), if you want to be precise. There is something wrong somewhere, because I have 2 "ranges" in my house and another 50m range in my sky garden, the values are pretty close converted, as in i can get 82k and 83k and 85k. For those variations it can easily be temperature of the pill, cell differences even with the same cell esp if it is unregulated etc.

Be careful of the word diffused as well, coz if you are talking about a soft box of 30cm x 30cm and measuring at 1m and 3m, then you will never get the converted values correct.....the 30cm x 30cm diffused is too close for 1m measurement. If 5m vs 10m comparison, then its ok. Everything is relative. Best is "point source".

22

Ok I give you another practical example which should be close to everybody's hearts. I have a Solarforce Masterpiece Pro-1 and the Fandyfire STL-V6 (same as Sky Ray STL-V2). Both get very close lux at 8.17m measurement distance, about 5% diff. And of course indoors the Fandyfire STL-V6 being a XM-L seems very much visibly brighter, even at 8.17m.

But on a 400m building, the very faint spot of the Solarforce MPP-1 is brighter than the Fanfyfire STL-V6, just very slightly brighter, though both are the same on the meter at 8m. This is the effect of focus, and apparently at 8 metres it still isn't enough. I am sure if i bring both lights up to the garden at 50m, it would tell me more. But heck i'll just take it as it is. :D

23

I thought of another possible thing, your light meter itself is not reporting accurate over the different intensities. There is nothing you can do about this. Just don't measure something at 15.3 lux and another at 300 lux to compare. But at 8m, you can get 500 lux for your Solarforce MPP-1 and 900 lux for your DBS aspheric XR-E, then ok. But if you are trying to do HIDs, then 1900 lux @ 13m may be pushing it. Try to go further.

It makes sense, even your DMM don't give the same accuracy at different ranges, esp if they cross the threshold of a particular range. Unless you have a precision DMM.

Ok, I used the Trustfire TF-3T6. 1300 lux x 3m x 3m = 11700

340 lux x 6m x 6m = 12240

1300 lux vs 340 lux, in the 2000 lux range.

24


It's probably important to delve more into why measurements at longer distance are relevant. This is related to the nature of the inverse square law itself, which assumes that the source of propagation is isotropic. I think this is generally true at reasonable distances for flashlights, but the point that it's calculated from is not necessarily obvious. A focusing reflector whose purpose is to create as bright of a spot at a given distance as possible might focus its beam ahead of the light, or even w/ perfect parabolic reflector the source would appear to be behind the reflector (whereas most people measure from the lens), plus it's possible it's not focused right, etc. By measuring at distance, it's a rough method of mitigating any such error. At one meter, a 10% error is only 10cm/4in, and 10% is probably more than most meters are off by.

25

Thanks. You've given me some things to think about and to try out. I'm sure after I use it for awhile and try out different light/distance combinations I'll see what works best and most consistently.

It's a fun "toy" to play with especially since I like aspherics.

Maybe you can go to my other thread entitled "Current question" and give me your take on what is happening there.

26

Ok, I just tried out the Trustfire TF-3T6, this is a pretty floody and "even" light so aiming the peak hotspot is not difficult. 1300 lux x 3m x 3m = 11700

340 lux x 6m x 6m = 12240

195 lux x 8.17 x 8.17 = 13016

Something like 10% diff between 11700 and 13016. It looks pretty bad, but it is not. Eg a 23.4k light and 26k light, throw pretty close.

1300 lux vs 340 lux, both in the 2000 lux range. Looks ok to me. I try to use this for my mid lights. I also use the same range with HIDs over 1 million lux, but i use them at 12.3m, not 3-6m range.

I think the accuracy of the meter for that particular range makes the most sense. I used the same 3T6 light and could never get it to work at 1m. It told me 15.5k with the 20k range selected. LOL!

PS. Diffused light like a piece of A4 paper in front of your XM-L, try not to do that too much, a lot of light may hit the floor/sides at 1m and not so much at 3m. Even if you are outside it may mess things somewhat.

27

Sounds good. The moral of the story seems to be to convert back to 1 meter simply for standardization but for the actual reading 1 meter makes no sense and a longer distance is required.

I'll do some more testing along those lines tomorrow.

28

I've got some chairs set up in the backyard now at 5 m and at 10 m. The light will be sitting on top of a flower pot on top of a patio table so the beam won't be blocked at all as it would laying on a flat surface like a table.

There will be no reflected light at the 5 m and 10 m testing spots and I think 5 and 10 meters should be far enough away to get similar numbers when converted back to 1 meter (we'll see).

I'll update this post later (after dark) if the numbers are consistent.

Update: HS-802 at 5 m 38,550 lux, at 10 m 30400 lux. Still not a close as I would like but in the ballpark and interesting.

I noticed that it's hard to get exact reading. I took several until they were close. This is just operator error. At 5 and 10 m it's easy not to hold the meter in the perfect spot although it shouldn't be hard as you can move the meter around and look for the higher numbers.

The closest reading I got at those distances was for a 30 mm aspheric where I got 17,750 lux and 16,000 lux but then I remember I had some tape on the edges of the lens (another experiment). I took that of and got 33,100 lux at 5 meter (which made more sense) and I need to redo the 10 m reading.

The numbers I got for the 38 mm in a HS-802 body were 40,675 lux and 35,500 lux.

The numbers for a 18 mm aspheric in a Ultrafire BJ08A body were 8,050 lux and (need to redo 10 m figures).

I think there is probably an optimal measuring distance for each light. Outside in my backyard it's hard to go more than 5 and 10 meter measurements. I could do greater distances but other objects tend to get in the way...bushes, trees, etc.

29

I guess this simply shows that for those throwers, their focus is farther than originally expected. Perhaps we really ought to be doing at least 10m, preferably 20-50m.

This also shows that for our usage, we don't need $500 NIST calibrated meters. It isn't even close to 10% accuracy for operator and enviroment induced discrepancies.

30

I agree, although if done in my backyard it isn't going to be much more than 20 meters :)

I'm happy with the accuracy of the meter however now that I've played around with it and see that the readings for flashlights (throwers in particular) need to be done at some distance. When you do that the numbers come into the range that seems reasonable from my readings from other sources.

My meter was just $25 or so delivered within 4 days. Most of the other meters have features that I just don't need anyway.

31


I did some basic calcs assuming the displaced isotropic point source hypothesis above. The [inelegant] eq is: (1-2*c+c^2)/(4-4*c+c^2) = ratio, where c is the location of displaced point source relative to 1m w/ vector coming out of the light as positive, and ratio is empirically measured lux at twice the distance (perfect is 1/4).

Using your numbers, ratio is 340/1300, and c solves to 0.0466, or 4-5cm inside of where you measured from (presumably lens).

To test this idea (by predicting 8.17m number), ratio of 3m to 8.17m should be instead 3.0466m to 8.2166m, or 8.2166^2/3.0466^2 = 7.273667007.

1300 /7.273667007

Ans = 178.726906084

1300 / (8.17^2/3^2)

Ans = 175.283787448

It's not 195, but at least it's closer...
32

I went out and tested 3 aspherics, 1 TIR and 1 "reflectored" thrower so everything but the TIR was very focused and the TIR was reasonably focused.

I tested each light at 5m, 10m, and 15m.

For the light that many have, the Uniquefire HS-802 (mine has a tailcap reading of 1A), I got very similar reading....33,925 at 5m, 33,000 at 10m, and 32,800 at 15.

The worse case (variance) was with the TIR. I think that's more operator error. At 15 m the beam is much more spread out and it's probably easy to miss the very hottest spot. The variance there was 18%. I'll do that over sometime just for practice to see what is needed to get less variance.

I tested 3 aspherics with the variance ranging between 4.5% to less than 1%. With these lights the 5 m figures are suspect (too high). After everything settles in at 10m and 15 m the variance is very little.

As far as actual numbers (just for information junkies) the Ultrafire BJ08A with 18mm aspheric was approximately 6,300 lux. A 30 mm aspheric I have was approximately 16,000 lux and the HS-802 with a 38mm aspheric was approximately 33,000 lux which was about the same as the reflectored model.

I think that's because of a very good reflector, maybe slightly larger than the aspheric, and a very crude aspheric (due to my grinding). I like it just because I like the narrow (no spill) beam.

I've learned that once you understand more about what you are testing and get a bit of practice using the meter as well that the results can be pretty consistent.

I've also learned that you can't really look at a beam and guess at the lux. I had one aspheric that threw well (but I don't have an area to check out its ultimate range) and I had a bit of tape on parts of the edge to get rid of some minor artifacts. When I started testing both with and without tape I saw that the tape was substantially reducing the lux numbers (and therefore took the tape off).

I also question the overall (low) numbers with the 30 mm aspheric. I'll retest that whole series as well. It was starting to rain and I had to hurry. It should be closer to the 38 mm aspheric than the numbers I got.

33

Yeah I think it could be pretty standard for most lights once you get past 10 over metres. But I think some really focused lights like big reflector short arcs, you will need really big distances, in the hundreds of metres. I think Ra measured his in kilometres, ie the exact working distance where you'd be using your light. It makes sense, lasers/coherent light obviously would not follow this rule.

This means i really need to go and measure my HIDs again. Maybe do it tonight with my 3 x DRYs. :D

34

Yes, I'm only talking about LED's. I think you have to measure most any LED at 10 meters or more or as you say...close to the environment you will be using them in. I'm sure with HID's you probably need more distance.

I also noticed that with less focused lights where you will be getting lower lux numbers you can take two measurements and be 10% off on paper and yet really almost have a match. My meter doesn't display fractions in the lowest range. In one case I got an 8 but if changed to a 9 it was a perfect match. At 8 it showed (on my spreadsheet) a 10% difference.

So when the numbers are so low that a single digit makes a 10% difference and you don't display fractions it's just a measurement/equipment issue.

I went out again and retested a few lights. Now if you only look at the 10m and 15 m readings I have less than 2% variance across the board.

I have one light that I don't understand why it's not registering higher but now I don't think it has anything to do with the readings. There is some reason it's not as bright as I thought it should be and I'll just have to investigate that more.

That's the fun of having a light meter (as well as a DMM) to be alerted to issues like this!

35

I also just figured out the 30 mm aspheric issue. I went out again to test and I saw that the light level varies (poor connection somewhere). When the connection is working properly I get the kind of numbers I was expecting (23,000 range).

Now I have those numbers at the 5 m and 15 m reading and not at the 10 m so that just tells me that the connection problem kicked in while I was taking the middle reading (10 m).

I also focused the aspheric so that the focus was perfect at a distance rather than on the walls of my house.

36

If you happen to read this thread again and if you have the chance you should measure your HS-802 outside at 10 m or 15 m where I found the readings settle down and post what you get.

I got low readings at 1 m but too high readings at 5 m. As mentioned the numbers settle down at 10-15 m. If your light is drawing 1.4A vs mine at 1A this should add no more than 5,000 lux at the most so yours should read in the 38,000 range.

Anyway it would be interesting (to me) if you are able to measure yours at that distance and post the results. I measured mine at night outside with no walls around.

37

Sure, I'm following this thread closely. The research you are doing is interesting.

I'll do what you propose tonight (if I survive the heat we have been blessed with this week). But from memory, on the first measurements I carried out, my readings leveled out somewhere around 5m. It's good that we both measure the peak value within the hotspot, so, at least on this detail, we have a better basis for comparison. Not shure if that is standard procedure, though.

38

I don't see how else one could do it or you would end up with numbers all over the place. If you're looking for throw that's how you would have to do it anyway.

It will be interesting to see your numbers. I'm liking the 15 m readings best as they are not only more consistent with the 10 m reading but they are easier to take the reading in the first place as there is less variance while you are taking the reading (easier to find the brightest part).

39

OK, here are the max values I'm getting from my HS-802, XR-E R2, driven at 1.4A. All measurements taken outdoors. Lux values converted to 1m. Battery at 4.20V, rested, before start.

5m distance ... 47.2k lux
7m distance ... 47.6k lux
10m distance ... 47.7k lux
20m distance ... 45.8k lux (perhaps a bit hurried, battery at 4.10V after 1 hour rest)

I agree that the difference between your values and mine is more than the difference in current draw (1.4 vs 1.0A) could explain. Could be the meters, but also differences between flashlight samples. With mine, the hotspot diameter is 10cm at 3m. In perfectly clear air, I can (just!) see a light-painted house at 500m appear from the darkness when I shine on it.

My previous measurement at 7m, done indoors, was almost 6k lux higher. Although I took some precautions to avoid reflections from the white walls, I completely forgot the glazed floor tiles (their dark color made me ignore them), and the sensor was placed rather low above ground.

This is the luxmeter I bought. That receptionist was the prettiest, and I thought that the Max feature might come handy:

http://www.ebay.com/itm/ws/eBayISAPI.dll?ViewItem&_trksid=p4340.l2557&rt=nc&nma=true&item=220796827173&si=ZaxL2aUFYf6eWNATmqNJMvEYy7A%253D&viewitem=&sspagename=ADME%3AB%3AEOIBSA%3AUS%3A1123

40

Thanks Agedbrier for taking the time to get those reading!

If anyone with this light and a light meter takes some measurements at a later date under these same conditions be sure to post in this thread. It would be interesting to see if the variance is greater in the individual lights or in the meters that are doing the testing.

My guess is that there is some level of variance between the lights with only a little of that coming from the meter if several tests are done under the conditions that we have done them in.

Agedbrier, it would be interesting if we had some other light in common and tested on of those just to see if you meters are close to each other or further apart.

Unfortunately, I don't have that many lights that seem to be common to a lot of people here. I do have a Trustfire Z1 angle light, a Zebralight H51f. I also have the 3 mode ITP A3 using XR-E or XP-E (not XP-G in other words). Let me know if you have any of those. I also have an older Surefire 6P incandescent.

We probably don't have a match on any of those but I thought I'd ask.