I’ve never seen any of them overestimated by such a massive margin.
All the test results on the TA Tube have been linear thus far except Charles’ X45 XHP70.2 hitting 17,000 lumens.
That’s the only result close to being right I’ve seen. Obviously if you calibrated your tube to a 1000 lumen Fenix light my TA tube will yield a similar result with the same light.
But if you Calibrated it to a 9000 lumen Fenix flashlight then your 1000 lumen Fenix flashlight would read stupidly high!!!
Right, I said that your X45 is doign something odd but you do not base a measuring system around a extreme measurement, you base it around the average.
Every light thus far has been well within reason of what it should be where we could get direct comparisons.
I have done TONS of testing with lights of all shapes and sizes and to put this mystery to rest once and for all I did a test with a special centering ring I made once.
I measured 3 separate S2+ lights and wrote down the numbers.
I then put all 3 of them in the tube at the same time with the special centering ring.
With the early designs that did not have the diffusers there was indeed a discrepancy in the readings.
After adding the diffusers all the lights read exactly what they should. Proving that the tube was now linear.
You own tests back this up as well.
The lights you are testing simply do not put out the lumens they claim.
The light lost / absorbed in the tube is a fixed percentage, it will be the same regardless of the lumen output.
Basically you’re saying every manufacturers, online reviewers/testers and my figures are wrong and only your Tube/Fenix is right?
I’ve stuck another 10 lights on this tube inc the proven Olight H2R and they’ve all read lower than any figure I’ve ever seen!!!
Technically I think it’s that all your high powered lights are under performing, and I think statistically this is unlikely. Guys, let’s see how we can solve this situation! The tube is calibrated specifically with that meter you’re using right now if I’m not mistaken.
But calibration is done using a 1000 lumen light right? But it’s a cheap meter, so is it possible that the meter (just your sample) is the problem and shows much lower values in the higher lux range? Is there a possibility that the tube is assembled incorrectly?
I am saying that this sphere is calibrated to an ANSI sphere and that if something is to be assumed to be off, the lights are the most likely culprit.
Light manufactures have everything to gain by having the lights read low.
On the other hand having the tube read low hurts peoples feelings and thus hurts the tube. Why would I want it to read low unless that is how it has to read to be accurate.
Lets break this down.
1: What are you wanting out of this tube? Are you wanting accurate numbers or numbers that match up with other numbers people post online regardless of how accurate they are?
It appears to be reading accurately right now. If you want numbers that lineup with online numbers, add 30% to the numbers.
2: Can you please post any Fenix lights you have? Yes, they are the only manufacture we have seen yet that consistently is close to the ratings. I don’t even own a fenix light myself but the Fenix lights people have posted so far have been within margin of error for the tubes.
If you really want to figure out what is going on then we need data that can be compared to others.
TA Calibrated his tube base on maukka integrating spheres’s DATA which I find his measurement seem to be about 5-20% less than other’s measurement from BLF, CPF & TLF.
Your MF01 cw 7,600 lumens is lines up with maukka’s measurement of MF01 Nichia 219c 7560lm@ 30 seconds.
You have to keep in mind that almost all of those lights are known for being over-rated.
The best comparison is fenix if he would post some fenix numbers.
As far as figuring out what is wrong, I am all for that but until data that is comparable to others is posted that is very difficult.
The meter has proven very consistent across all the other spheres and I have not had any issues with my 3 meters reading off at higher outputs.
I even compared them to a $150 EX300 (I think, it was borrowed) meter and they read within margin of error of even that meter at LUX readings of many many times what it sees in the tube.
You have to remember that the tube is only reading maybe 10k lux, almost nothing compared to taking throw readings with 20-50k+ lux.
It is of course always possible that the meter has an issue but his own ceiling bounce numbers that he says are trust worthy prove that the tube is actually reading better then the ceiling bounce numbers.
This shows that the tube is in fact NOT reading lower as lumens go up. The light is simply not putting out the lumens it should.
To the contrary, the Sphere actually reads HIGHER as lumens go up.
All of these things put together would say that the tube is not the problem.
Until some fenix numbers are posted we just can’t really figure out much more since we can’t directly compare the numbers reliability.
I can’t figure out why he won’t post fenix numbers. He said he has some Fenix lights.
Did you also test every sent out tube at higher ouputs of ANSI 5000+ lumen? Or did you just calibrate it to a ~1000 lumen ANSI source? This question is related to my previous post.
Edit:
I was under the impression that the ceiling bounce test was done using his phone, and not the supplied TA lux meter: I could be wrong of course.
Good find, further confirmation that the numbers are reading well within spec.
I am very happy with that, a mere 0.5% off of a real ANSI sphere.
That is far better then I would of expected, could be luck but still, I am extremely happy with that.
Yes, I also tossed a 5-10k lumen light on them for a quick confirmation during calibration but since these lights are not super consistent they were just used as a confirmation.
That said All of the tubes read the high output lights correctly, nothing was off by 50% for sure. That I would of noticed.
The biggest reason I am pretty sure the tube is fine is that everytime we get a comparison of a light to someone else, the lumens are well within margin of error.
I also understand that manufacturers’s lumen ratings are often overrated. But Maukka’s measurements seem very reliable. Just look at Maukka’s measurements of the X45, X80, M43 219C (IIRC). From what I know KG’s high output lights all seem to be outputting a lot less than what Maukka measured. So it’s not about manufacturer’s spec, but Maukka’s spec and from other reviewers (not one) where output is converging to a certain figure that is contradicting KG’s measurements.
Of course, there is always a possibility that all KG’s lights are under performing, but like I said I think statistically it’s unlikely.
KG:
I believe the tube is working exactly as designed. In fact better then designed, I am seriously impressed with how close the numbers are to others.
That said I do not like to think that someone is unhappy with something from me and I want to make it right.
Return shipping is not really an option sadly (It cost me $65 just to ship it to you).
So as a middle ground, how about a $50 refund?
Would that be acceptable to you?
Yes… TA tube is perfect. I measured my fenix tk15, 1000 lumen light again. I use lg mj1 10 amp cell which fenix would be used In testing… I got 948 lumen @ 30 seconds…
At this point… I can’t figure that out either??? :person_facepalming:
The X45 I can not explain, it should be ignored since it is an odd one out.
The biggest issue with these lights is that they are not proven consistent so comparing them is difficult.
The M43 for example has at least 2 different driver versions and different outputs to match IIRC.
Also tint has to be factored in, this meter will indeed show a tint bias compared to Maukka’s.
Some fenix lights would put a lot of this to rest if he would post numbers for them. Those could be much better compared and the problem figured out. Without numbers that can be accurately compared though, it is impossible to figure out what or if there is an issue.
I know on my sphere it compares very very well with an official ANSI test of the MT09R at 13.5K lumens. So the design is not biased at high outputs and nothing yet has shown that there is any kind of bias as lumen output goes up.
Just that the high output lights (aka, the ones that are hardest to reach the rated numbers) are not reaching the numbers they state.
Another thing that needs to be factored into some of these lights is that the latest flip chip LED’s do have a break in period and they will drop output as they age.
We don’t have exact numbers on this but the running estimate from those of us that have noticed it is around 10%+ over an unknown number of hours.
So if all of his lights have been used quite a bit for reviews ect, then that can easily explain a fair amount of the lost output.
I noticed this earlier but forgot to mention it, did you loose the ring that holds the centering ring in place? I can print off another one if so, I know it makes it a lot easier to line everything up.
No… i have it… i am constantly testing all the lights. Its get on my way. Lol… thanks TA…
If KG accepted, he could then sell his lumen tube at a lower price as “used”, probably recoup all his money and be done with the whole stuation.
Bingo! We have a winner!
There are only a couple online reviewers who have professional integrating spheres because those setups can cost $10,000 dollars or more. Maukka has a nice setup. I think Koef3 either owns one or has access to one. That’s the only 2 people I know of.
Manufacturers want to sell lights so they tend to exaggerate the output. Some exaggerate a little and some a lot. I think Fenix does not exaggerate (much).
The vast majority of online reviewers/testers use homemade measuring devices that are typically calibrated based on an average of many factory lights rated specs. This is why they tend to get higher than real numbers. Even getting ahold of a calibrating light that has been properly measured has been almost impossible or else prohibitively expensive (until recently).
Since you have so many large and powerful lights, maybe you could build a really big homemade integrating sphere (maybe with a large ice chest or something) then buy a calibrated light source from Maukka to figure out the correction factor. Use a lux meter of your choice. Maybe then you will get readings more to your expectations?
I have some of the same lights as KG_Tuning, the readings all look reasonable to me. Here are readings for some of my lights in the TA Tube. The last one on the list, the HDS Ra Executive, was factory calibrated to 100 lumens.
Astrolux MF01 7800
Noctigon Meteor M43 6080
BLF Q8 4200
Emisar D4 XPL-Hi 3520
Sofirn C8F 2750
Acebeam EC50 Gen II 2420
Emisar D1 XPL-Hi 1292
Emisar D1S XPL-Hi 1296
BLF A6 1300
Zebralight SC62W 977
Jaxman E2L 811
HDS Ra Exec Hi CRI 100