milk carton Tetrapak box is what selfbuilt uses for smaller lights. for bigger lights he uses ceiling bounce method because the milk box is too small.
yes, any multi-mode Fenix light is helpful. a single Fenix flashlight model would be sufficient. As confirmed by the integrating sphere, all Fenix lights and their modes are on the same lumens curve in an integrating environment. ceiling bounce method for example, is no integrating environment. and measuring lux isn't either. note that UPz's sphere does not measure lux but miliVolts of a clever electronic circuit with a clearly specified light sensor by Sharp Corporation.
milk carton method is totally cool, as long as you have enough Fenix lights to verify your lumens estimations based on a luxmeter reading.
I don’t think much accuracy can be obtained with most home-built IS devices unless an involved/expensive calibration process is undertaken. The main benefit of these tests, if done consistently, is a relative measurement of the lights tested with your own device.
The results can be usefully compared with those of another device if both devices have been calibrated against the same standard, as Match suggested in his IS construction thread a couple years ago.
I use my wife as my lumen testing device. She is as accurate as my cheap meter. I do a ceiling bounce in a dark room with her not seeing either light. Honey which one is brighter number one or number two? Is it terribly accurate? Nope but she can tell that my terminator is putting out more light than my SRK and that my SP03 is not, etc. etc. Anytime I post any kind of numbers I say it’s just a guess based on brightness compared to lights that many people have a decent idea of. For throw it’s kind of the same. I have various targets that some lights can reach and others that they can’t. It’s cool to know that if someone is standing a half mile away that they might see a tree dimly lit up next to them by my light. I have used my wife for this as well only I go for the walk and have her shine the lights while I stand next to the trees. Again this doesn’t give me numbers but I think it’s as accurate as I can get without spending a boatload of money.
I keep saying that one of these months I will spend my budget on testing equipment but then I see how nobody seems to be all that accurate anyway and my desire for new emitters or hosts wins out and I end up waiting for parts for my next build instead.
Maybe next month I will spend my budget on testing equipment. 0:)
Me and rdrfronty/manxbuggy1 have the PVC pipe light box, which I think TurboBB has as well. Once you get a good feel for the calibration, it seems pretty consistent from the 100 to 5,000+ lumens lights. We just use different size hole, dark gray foam adapters for fitting the flashlight head into on one end, and the other end is where we velcor mount the light sensor of the meter.
The purpose of a standard is to reduce variability by providing a single point of reference, such as one flashlight. I think results would be more comparable and repeatable if a single copy of a Fenix flashlight was selected for the standard. Using a single model would be less effective but perhaps more practical. Using an entire brand of flashlights as a reference point would seem to defeat the purpose.
my main point was that Fenix data lie on the same curve (data analysis in a x-y-coordinate system) whereas others <xxxfire, thrunite, nitcore, klarus> don't. the latter group produces inconsistent data, the former consistent. and due to the consistency, it's no big deal to take just 1 identical Fenix or several Fenix flashlight models as reference point because interpolation and extrapolation is valid.
it's impossible to get a single matching curve (data analysis in a x-y-coordinate system) with a set of Nitcore lights. so if you take Nitcore flashlights as reference point, then i agree, everybody *must* use the same identical 1 single Nitcore flashlight model.
Nitcore, klarus and thrunite lumens scales s*ck so bad, one cannot believe it omfg. They are the worst among premium china made lights. Don't you give a **** on what they publish as ANSI lumens claims pah! :Sp
Just searched my photo bucket lib and don't have any. I'll have to take some and post. Think manxbuggy1 may have posted some, but not sure -- he built mine.
I don’t have much confidence in the absolute lumens claims of any of the manufacurers named. I suspect that independant testing by a reputable lab would show enough variance from published specs to make many of their claims laughable. But that’s not my point.
If your objective is to make practical measurements that are useful to others then manufacturer’s claims aren’t very important. Just choose a light for reference purposes and pass it around. If you’re using more than one copy of the light as the standard then manufacturing tolerances/consistency becomes a factor and your standard is less reliable.
Here's a direct quote pm from rdrfronty (hope he doesn't mind):
"Basically our lightbox is a PVC one. Made out of 4” p-traps, with additional 90’s and caps, and a glass platform. The p-trap has the extra 90 stuck on one end and rotated up to accept the lights. A glass platform is siliconed into that 90 to allow the lights to sit on it. On the opposite side a cap was put on to seal off the box. I drilled a hole just large enough for the meter sensor to fit into. I then have a Velcro strap to hold said sensor in place. Now on the light receiving end we have multiple sized foam rings that we put around the bezel of the lights to help seal the light from coming out.
We also went through the ptrap, elbows, caps, etc, to scrap any bad mold lines, sharp edges, etc. that light could catch on. So the end result is a pretty smooth curved tube with no flats walls & no corners, but also no direct light beam being allowed to touch sensor.
So with the sensor strapped on and appropriate meter levels selected, the light sitting on the glass shelf, donut surrounding the bezel, we turn the light on, start timing, record start up numbers, and then finally record the final otf numbers once 30sec. Is reached. Now the thing is these PVC meters, and almost all homemade lightbox of any style, have a different volume that “real” light boxes. So the numbers you get will be off. So the trick is to get as many and as accurate of ANSI rated lights tested on the lightbox to create a baseline number to multiply or divide by. We’ve tested a good 30 ANSI rated lights through out the lumen range to help get accurate baselines. And with this setup that baseline number actually works well from my 25L Preon 0 up to the big boys. Some lightboxs will need a couple baseline numbers, a high and a low range. But ours worked really well accoss the full range."
There's a reason why an IS is minimum of $10,000. What we do has nothing to do with true readings or even ANSI lumens output. All of the home made IS spheres are subjective and only meant for something for your reference, but they would not be true lumens, nor would they be necessarily repeatable with a whole span of different lights.
In other words, any readings from a home made sphere, or worse, (light box or PVC), are just speculative and cannot be considered true lumens readings. I think too many people take some of the results they see as gospel and they can be totally way off from true lumens readings.
I would no way, no how claim my measurements are accurate to true lumens, but I also know it's damn close, in the neighborhood, shows consistency over and over again when done and setup correctly. I see lots of consistency from unit to unit (many TN31's and many Shockers for example). Also I've done many battery tests in various well regulated lights and poorly regulated lights and can easily track differences in lumens output as a result. Best part is making a tweek and seeing the measured result, and also, getting someone else with a totally different home made lightbox, calibrated completely different and getting the same results you get - amazing coincidence that seems to happen quite often...
