Yep, even 1mm will have a noticeable effect. The same goes for the 1st piece of PVC as well, how hard it is pressed in can have a surprisingly noticeable effect on the readings.
Now factor that by the 4 joints in the sphere (which you had to take apart to make any changes and then reassemble to measure and hope you got it together with the same force every time, which you never did) and you start to see the tip of the iceberg these were to calibrate lol.
Sounds great!
Missed this before, I have not tried much yet, until I have a more exact correction factor to aim for it is hard to know what will work.
I am thinking some form of DC-Fix type sheet is what I would like to use, unless I happen to find something that works perfectly. I am going to check home depot sometime to see what kind of options they have. Some very light window tint could work but I have a feeling it would drop the readings too much.
I also cut some discs out of the left over/broken diffusing sheets I used in the spheres to see if they happen to drop the readings by the right amount.
So i tested my fenix tk35ue… there is a guy name Kospap who tested the fenix tk35ue and we got similar result… I got 3486 lumen @30 seconds… below is his graph… look like my TA tube is working great with .70 factor.
How does Kospap measure it, does he have an integrating sphere like Maukka?
Edit: I see his review. He says he confirmed the lumen output, but he gives no details on how he did it.
Here is just a few.
all readings on original tube at 10 seconds
Olight S1 rated 500 aluminum side down 744 grey side down 710 AW16340 finishing charge 4.08
Zebralight SC62W 930 1245 1175 zebralight battery 4.10
Nitecore TM16GTVn 6300 6650 6540 unprotected 30Q 4.05
Thorefire VG10 847 1178 1050 30Q 4.02
Thorefire BLFQ8 5500 5880 5680 30Q 4.04
Niwalker mm18mbvn 8500 8480 8380 30q 4.05
Emisar D4 3826 4550 4360 30q 4.02
Emisar D1 1434 1828 1747 30q 4.04
4-7s Malestrom mmu-x3 1600 2520 2480 four sevens 4.01
I have no clue… i was reading it on cpf and it say light box… . we would have to send PM and asked him how he calibrated and stuff.
I sent PM to him… i will let you know.
We always put the reflective tape side down into the tube. This is the way TA calibrated it.
If you flip the spacer over to have the painted side down, you get unknown numbers.
Yes, the silver paint was an experiment to try to make things simpler but it didn’t work. Too bad, that reflective tape was way way more work then it seems.
I spent last night reading through all the posts - interesting stuff!
Curious, does ambient light impact the measurements at all? In other words, does it matter if the measurements are taken in a dark room vs outside in the sun?
The PVC is thick enough that the most I have ever really seen is 1-2 lumens from ambient light through the pipe itself.
So yea. Maukka lumen are legit… he tested astrolux mf02 cw @ 30 seconds and got 2109 lumen…
I tested my mf02 nw @30 seconds and got 2044 lumen.
I am using .70 correction factor as worst case scenario…
Yep, I kinda figured they would be…. since he does have a real Integrated Sphere. 
I think it’s only fair to mention the MF02 has had at least three different driver versions with three different turbo output levels. When the MF02 first came out, turbo was about 1.8 amp. Then it got bumped up to about 2.2 amp. The very latest versions within the last month seem to have been bumped up to about 2.7 amp.
Mine is the latest version… we can say for sure mf02 are no more than 2100 lumen…
I just had some time to play with the lumen tube very briefly. It is VERY easy to use compared to my ceiling bounce and calculator method. It is also very consistent regardless of the flashlight head insertion depth, which is a sigh of relief because that was something I was worried about. The measurements are very high though as already confirmed by others. I get about 27.5k startup lumens on my MT09RTA whereas using my previous ceiling bounce standard, I was getting 22.8k startup lumens. Once TA gets the calibration corrected, this will be a fantastic device. I don’t regret my purchase a tiny bit I can tell it required alot of manual labor to construct this along with the discs.
Now I will have to store this away until I find a place to put this in my house. I’m out of room already. My room is already littered with flashslights and batteries.
Thats good bro… i calibrated my meter to fenix standard, and took the .70 factor lowest number for the worst case scenario.
Haikelite mt09r TA triple xhp 35 cw. 5425 lumen @ 30 seconds
Haikelite mt09r TA tiple xhp70.2 cw. 18,200 lumen @30 seconds.
Haikelite mt07s. TA. Single xhp70.2 cw. 6552 lumen @ 30 seconds.
Sorry if it’s too off topic, but is there an easy way to tell what version my MF02 is? Thanks!
When you bought it would give you a general idea. The only way for sure is to measure the amp draw on the emitter.