Lumen and lux measurements, why cant we all try to be on the same page?

I actually followed that link. It’s a surprisingly interesting read! Thanks! I’m not sure we can really apply that last piece of advice - to get off the standardization effort - in the context of this conversation. In fact, it doesn’t fit at all. That advice was given to a person seeking to become an expert in their chosen endeavor. We’re not talking about becoming experts here. An expert may very well be expected to balk at the idea of being merely comparable to others. But, we’re talking about a place where sharing information about our chosen subject is the very reason we gather. Making the information more meaningful by standardizing to a point of repeatability would be the higher goal. In this case, excellence requires being relevant.

As has been said before, the measurements we take are only going to be good enough to compare light “A” with light “B”. But, shouldn’t we have a reasonable expectation that one person’s light “A” will shine as brightly as another’s? If the readings on light “A” can vary by as much as 20% from person to person, then how do we compare it to light “B” reliably without each person having to possess each light? With thousands of lights in the world, can anyone afford to own them all? In most studies, polls, elections, and other such collections of data, the outliers are usually considered irrelevant. They aren’t given consideration. Nobody cares if the one guy whose measurement is 30% higher or lower is more accurate or less so. They go with the consensus. Don’t we do the same thing here - try to reduce spill and make a tighter beam? Even when trying to make a flood light, is it not more desirable to make it somewhat uniform? Does anyone intentionally make a light with 3 emitters aimed in three directions, and not rather aimed in unison at one target?

I hear you guys that are saying that dead-on accuracy is impossible because of so many irregularities in the flashlights themselves. I know what you’re saying, and you’re right. But, that doesn’t mean we can’t try to get as close as possible. And it doesn’t need to be as hard as some people suggest. For instance, if one person’s measurements really are reliably ~20% higher than everyone else’s, then all we need is a correction factor for that person’s numbers. Let’s say we have 10 people testing lights and their numbers vary from person to person typically by the same amount each time. We take that amount as a correction factor and apply it so that their numbers are closer in line with each other for the “same” lights. Then, when we see differences, we call those actual differences in manufacture from one light to another.

DavidEF;

Fine but who is going to skin the cat? Unless and until a large number of THE SAME LIGHTS are tested by all the people with Lumens spheres within a relatively short period of time with the same batteries charged on the same charger and then all the data is processed from each person and a correction factor assigned coming up with common readings is not going to happen. The logistics are a killer and this is the type of thing that needs to be repeated regularly as measuring equipment ages etc. That is why in industry there are calibration companies and labs set up to do periodic calibration of test and inspection equipment, and even a lot of production equipment including torque wrenches. In industry if it measures anything it has a calibration sticker on it.

Based on the evidence cited by people who have done the measurements even different manufacturers cannot agree on Lumens out for a given light so to expect a bunch of hobbyists to do better is asking the impossible it appears to me.

This thread has been an education and I would like it to be a sticky as it is so educational but I also think that the subject has been discussed to death and should be closed. Unfortunately those who want perfection will never get it just as even manufacturers have not done so based on info regarding test results presented here. Discussing it ad nauseum is not going to change the situation that currently exists.

I think the reason Norvig suggested getting involved in standardization is to better understand the things which matter to other people, and why and how deeply they care. It’s not about actually standardizing anything. And the reason to stop is to keep one’s sanity… to keep this from happening:

Take soda, for example. Er, no. That’s too complex. Take cola, for example. Or just Coke.

At the Coke factory, the product is pretty consistent. But by the time it gets into a customer’s mouth, the taste and formula vary wildly. The syrup they produce has a limited shelf life, so the taste changes by time since manufacture. It goes into a bunch of different types of soda machines which all behave a little differently, and may or may not have been cleaned recently. The soda machine must be calibrated and maintained by a person, and nobody is quite the same as anyone else. The machine mixes varying amounts of water and CO2 in with the syrup, based on the machine itself, the charge level of the CO2 container, the amount of syrup left, etc. The local water source changes the flavor too. CO2 retention varies greatly based on temperature of the liquid, and the overall taste changes quite a bit with temperature too. A bunch of CO2 is lost immediately based on the velocity of the liquid when it lands in the cup. People add different amounts of ice. The ambient temperature changes the rate of decay once the product is dispensed, and it doesn’t taste the same after it has been sitting for half an hour. And so on.

I bet Coke would really love to make sure each person gets exactly the same beverage experience, and probably goes to great lengths to standardize as much as possible. But how often do people complain that their drink’s CO2 level is off by 20%? How often do they even notice? There’s a point where one must either decide it’s “good enough” or leave sanity behind. Diminishing returns transforms the effort into something of an entirely different nature, a matter of philosophy rather than practical application.

I mean, if you think there are a couple of restaurants where the Coke really doesn’t taste right, then definitely let them know. But try to keep in mind that the goal is better-tasting soda, not convincing them to stop serving beverages.

for me, leaving sanity behind is inevitable in this hobby and it sounds like an attractive option :bigsmile:

And in much of the world Coke is still made with genuine Sugar as the sweetener. IIRC Coke syrup for export does not have the sweetener added so it is cheaper to ship due to greatly reduced bulk. In Nevada many stores carry Mexican coke which is still in the classic glass bottles and is sugar sweetened. Very different tasting than all currently USA made Coke that I am aware of. Sells for a premium price too.

Well, okay. There is that. :slight_smile:

I mean, have you seen the curios full of lights? Practical purposes stopped being relevant like a hundred lights ago. :stuck_out_tongue_winking_eye:

Yup, I have seen those, and where else can you show them off without being laughed at but here? :-) (ok, I can't suppress a mild grin ;-) )

BTW, the unhealthy direction this hobby has taken in my case is not so much showing off my vast collection (which is not that vast comparably, it all still fits in a shoebox, although barely), it is the totally unneccessary drive to measure light correctly (a quite hopeless undertaking), exactly what this thread is about. I recently spent 6 times as much money on a decent luxmeter than I have spend on my most expensive flashlight

@DavidEF

I just have to say one thing. YES! :)


Toykeeper and Djozz.

I was sane once, then I joined BLF.

I can certainly see those curios being useful. My lights are floating all over the place and I dont want to put them all away. :D

Djozz, keep us updated on your "adventures" with your luxmeter. Im curious how consistent it will read compared to your LX-1010B on different wavelengths. :)

I did some testing in this thread :-)

Ok, I am having a bit of trouble creating a baseline. I don’t exactly understand how it’s done. Let’s take the Fenix LD15 for example. It’s low mode is 8 lumens, I get approximately 29.7 from my luxmeter… I divide 8 by 29.7 and get .2693 which is my multiplier. Ok, so now that I have one multiplier…. let’s say I have three or four other ansi/FL1 lights. What do I do then? Get a multiplier from each light and then figure an average?

I’d say toss out the high and low values and average the rest…

You can either try to come up with a single multiplier to apply to everything you measure… or you could try to use a curve-fitting function. Basically, plot a bunch of points for a bunch of different lights, calculate a curve which goes approximately through the center of the point cloud, and use that curve instead of a multiplier.

I’ve been using the simple multiplier option, as have most people around here. I used some Zebralights and selfbuilt’s measurements for the initial calibration, compared it to some other figures I found around BLF and other sites, and found that the results were very close to what other people were getting. So, I’ve just called it “close enough” since then.

If I ever build that sphere I’ve been meaning to make, I’ll probably also go through the trouble of making a curve-fitting program. Until then though, a milk box and a multiplier get me close enough.

zunzun,com … the mother of all curve fitters. It fits your data to about 40,000 different equations and ranks the results. Can also generate code to implement the curve fit.

Frankly, curve fitting is not the way to go about it. Once you feed a curve fit routine a point outside of the “training” data used to generate the curve fitting equation, very bad things tend to happen. Also, to make curve fitting worthwhile you would need a BUNCH of different well characterized lights of various output levels, beam shapes, color temperatures, etc and a way for your sphere to measure those. I’d go with a simple single calibration constant.

Hmmm… I don’t know if I’m more or less confused. Is a curve an average? Ive been using a single multiplier for a while because it’s gotten me close on several lights, but others seem a bit off.

I can’t tell if the consensus is to find a single multiplier or use several. Toykeeper and texaspyro… Do you use a single consistent value as your multiplier or not?

I use a single number, in fact I use a single flashlight (not an average of several) of which I know it delivers a very constant output on one of its settings, and keep this flashlight solely for calibration purposes (a bit of a pity because it is my SWM D40A that I would love to use for camping, I use the high setting -550 lumen according to SWM-, not the turbo or lowest setting because they tend to vary more). If this number is off from the real lumen, at least all my measurements are off with the same factor and can still be compared to each other, and if I finally find the ultimate calibration, all my measurements in the past can be simply corrected with a single correction factor.

Yes, a single consistent value as the multiplier.

Because it’s easy, and because a straight line is a good enough approximation of the actual curve we’re trying to measure. We’re not doing rocket science. :slight_smile:

I have this light too. I use it to find my candela rating on other lights because it seems a pretty consistent light from all the reviews. The recent G25c soured me on Sunwaymans output figures though. That light was way off on all specs. What happened?

I recently got a Zebralight sc600 mkII… Donated by a very kind redditor. I may try my hand with this light to see where I end up. Also gearbests olight sales have found me in the possession of two of their lights now, so we’ll see where we end up with them.

Ok thank you, some times I need thinks explained as if I’m 5. So you use a single mode on a single light like djozz does?

At least two points are needed to define a line, but it might be safe to assume one of those points is at zero.

How I started was by measuring each mode on a ZL SC52. Given those raw meter values, I adjusted the multiplier until the readings were all as close as possible. Everything between 3 lumens and 300 lumens lined up almost exactly, but above and below were off.

Then I checked it against other lights with known consistent output, mostly based on reviews by selfbuilt, and adjusted a bit more until I got the overall set of lights pretty close to his numbers.

But in the end, I’m using a single number.

Perfect, thanks! Exactly the info I was looking for.