Hey guys, @Daninho expressed some interest in a PM of possibly sending a sphere and flashlight to a lab that could nail the higher (1000 lumen) calibration. I think that would be awesome, and perhaps we can all chip in and help with that bill. I think it would be awesome and I would be willing to help pay for this too if it’s more than just me and daninho splitting the bill.
Edit: I’m for about $75. He estimated it at $200.
Hi,
no i wrote a mail to a so called light-lab and they are able to measure flashlights, not a sphere. They just measure how much lumens a flashlight can produce which will help me alot since i dont know how accurate the manufacturer lumens ratings are. This is the first step to more accurate lumen reading because i can imagine that this sphere was calibrated with a Fenix flashlight based on manufacturer claims which are good in case of fenix but maybe not perfect.
There are other manufacturers with “honest” lumen ratings like surefire or mag lite. I feel this is the only way to really check the performace of the sphere.
I tested a nitecore MH20 in the sphere and i always get like 8xx lumens from it in turbo and the rest is also a bit far off, so i dont know if that is because of the sphere or because of wrong factory claims.
I usually have to select the 0.0050 factor to get the lower lumen values in range to the factory ratings but then the turbo and high is to low, hmm.
When you say more then 1000 lumens, is that using an NR filter?
Well I meant at the top end without a filter, which is near 1000. We can verify NR filter performance without the lab. I guess we need to talk about it more, but I understood they were willing to take a light and the sphere and go back and forth until they read the same.
So basically you just need the true lumen readings of a flashlight, not the claimed ones. Then i have a new multiplier for that sphere because nobody knows how accurate Joshs old Fenix light were in terms of lumen for getting the multipliers in the app, i hope that made sense.
So, for example, the procedure would be
It wouldn’t matter what the output really was or even if it dropped off. Of course you will know at the moment it’s in the lab sphere, but what it is doesn’t matter.
When you know for sure that for example the Olight S30 produce lets say 567.8 lumens on high according to the laboratory its all good. You can use this light as your reference in your sphere from now on, just write down the readings of your luxmeter, like the same thing you did with your fenix reference light in the beginning. The only difference is this time you know exactly how much lumens the light will produce because its lab tested. The Fenix light has only factory claimed lumens, nobody knows which battery they used and every Fenix light is different, even when its the same model.
The Fenix has some special regulation in it you don’t find in others to my knowledge, and it does not get hot. An Olight, for example, will output very, very different lumens depending on the exact battery charge, the time draining the battery, and even the temperature of the LED at the moment. Give it a try in the sphere, the numbers drop fast and keep dropping.
I think only on the higher modes and yes the battery needs to be full everytime and they must use a stopwatch and measure after 30 secs or so. I dont think the lower modes will drop fast, maybe a light with a bigger head is needed, that can run on 400 lumen for a long time. The turbo modes dropping fast, i noticed that.
Well if the exact details could be worked out on how this would work and what light would be used then I could toss in a few bucks (just don’t have much to spare but also curious how accurate the readings are).
I will say that my 10 year old fenix’s are the most consistent lights I own by far, seems that no matter what the battery voltage (within reason of course), or how hot it is, it always puts out 139-141 lumens. Even after all the abuse it has been through.
So a good fenix is ok with me as a calibration light.
The more i think about it and read about it in other forums the more doubts i have. Fenix flashlights regulate well, the Fenix LD41 could be good with a mid mode of 400 lumen. The problem is that my sphere is probably different to yours even when it appears to be the same.
My olight S15 lumen readings should be closer to the factory lumen ratings. I have calibration number 520 and i think i get the closest results with the 0.0050 multiplier but its still a bit far off. For example im not able to measure lights with moonlight, i get for example 2.8 lux which equals 0.0 lumens according to the app. So i think something is a bit wrong with my sphere, thats why i want a calibrated light, maybe i need a new multiplier and base my future calculations on this light.
I own mostly small one celled flashlights like olight s30, s15, klarus mi7, nitecore mh20 cw, on the road m6 …etc… Maybe you have one as well for a test in your sphere? Or maybe the other owners of that sphere can test it as well?
It seems to be generally agreed that Fenix has some very consistent lights, even across many different manufacturing batches through time. But there is still some variation so at best choosing a Fenix (or other) light to use will not result in exactness for all. Each sphere/meter combo will have its own minor variations. And a light tested and calibrated at a given level may not truly reflect readings taken at other levels. So while any individual could have a few quality lights tested and use them to calibrate their own sphere, that will not relate to anyone else’s sphere except in a general way.
So we’re back to the original concept of use, that this is more a comparative tool for use by and between us flashaholics than it is a laboratory instrument giving exact readings.
I sorely wish someone would build an economical regulated light tested and adjusted to a standard level of output which could be powered by wall current adapters world-wide. By having a few of those made at different output levels everyone could have a reasonably solid basis to work with. Not exactly an ANSI standard light but close enough for what we do, say +/- 10 lumens at high output levels and within a couple lumens near the bottom of measurability. Those could be sent around among us continentally at low cost so that a few such sets would get everybody on the same page economically while giving close-enough accuracy for all but the most demanding of testing. I think this idea far more likely to assist us than by having one or two flashlights calibrated then hope that similar ‘identical’ lights will coincide.
Phil
Correct, this is what I was getting at before in this thread. This sphere does exactly what it is supposed to do, it reads consistently.
Consistency > Accturacy
By being consistent I can measure any light I want against any other light I want and frankly that is all that matters. By us all having the same spheres we can also compares numbers from anyone’s sphere with any other one.
For those in the automotive world this is like the dynojet of the car world. The dynojet is the only comparable dyno on the market, others like the mustang dyno are not consistent and vary wildly between shops and even different days. So there numbers are useless when comparing numbers online.
When people really want to brag about their numbers online, they get on a dynojet, that way everyone knows how it compares to their own cars.
Everyone also knows that dynojets read “high” compared to the actual definition of HP but who cares as long as it is consistant?
Here’s the 3D printing files for all 11 adapter sizes I drew, in case anyone wants to make use of them.
I just got PM’d a question about if I adjusted the lux meters in any way. Here’s the answer.
I didn’t adjust the meter’s factory calibration, I’m not even sure you can. What I did was test an exact spot in my basement bathroom with all the meters, one at a time. I put the meter reading on the back of the unit. The average of all the readings was 555. So if your meter says 560, then it reads a little brighter than it should. If it says 540, then it reads a bit dim. This percentage of error becomes the ‘meter calibration’ in the app. This is how it uses math to make them all agree with each other.
how accurate can the meter be on pwm lights?
any specs on rise-fall response of its sensor?
When measuring moon mode on a light with PWM the lux reading is rock solid. If there was an issue I would expect the readings to bounce around each time the screen updated. That’s all I know.
The 1330 meter specs state: “Sampling rate: 2.0 times/sec” so PWM is not really a concern.
PS. I own one and think it is great!
I am interested… thinking about where I would put it, why it is important to me…. before going any further, am I too late for this?