Yea me too. TA lumen tube works really great.
Thanks for sharing steel!
Thanks for the results. So neutral high CRI is spot on, cool white lower CRI is about 3% too high.
I see you have two extra diffusers there. How many did you use on the TA tube?
Here we go with the math again.(Steel’s lights)I took Maukkas readings at 30 sec.
So we got 35 and 34.1 which is somewhere around .972. (so if 40 is the reading-38.88)
And we got 98 and 94.5 which is somewhere around .965. (so if the reading is 485-468.02)
Now, since we have a difference should we average the .972 and .965? ( .968.5 )
We get up into the higher numbers with the S2 and the number goes smaller ,say .945 do we just average all numbers or just the two calibrated lights we have?
Sorry to be so dumb about this. Math is not my forte’.
Since the lux meter is probably not filtered perfectly I would use different correction factors for different types of lights. Check out djozz’s great lux meter tests for more info. Test1, test2, test3, test4, test5.
The difference is mostly due to the tint of LED I am guessing.
These cheap meters are effected by the tint of the LED. Meters that are not biased would cost as much as the whole sphere.
There is not much that can be done to correct for the tint bias sadly outside of buying a $100 lux meter.
Yep, I agree. I know for a fact that these meters (and every other meter I have tried in the sub $100 price range) are biased to cooler tints.
I just yesterday found out that my most used integrating sphere warms up the spectrum by a few hundred K, maybe I should use a cheaper luxmeter to compensate for that :party:
I would not do any math. It’s not about super accuracy. I would use the number right off the meter for NW lights.
For CW I would just bump it down about 3% worth. So a cool white light measuring 1000 lm would be 970, etc… Easy peasy.
I have a new Extech LT45 luxmeter (to be tested together with some other new luxmeters) that because of the imperfect spectrum sensitivity has pre-sets for different light tints, which is about the same what you suggest here. 
Smooth surface, face to face
Grooves are outward
Scotch tape fixed on both sides
Minor reduction of light
Put into the tube
Leave some gap between the illuminometer and the tube
Fully charged battery eneloop AAA
I get perfect data
BLF 348
Olight i3E EOS
I can see a lot of light escaping.
Maybe those small lights need a tighter fitting centering ring?
Not if it’s calibrated correct right now. Don’t touch anything, just enjoy 
I disagree. Plus I don’t think it’s calibrated. If keeping the leaking light in the tube makes the readings go up, then it’s reading high.
In my opinion, the fit around the light should be a bit tighter (although it doesn’t always seem to make that big of a change) and the lux meter needs to be pushed all the way in. This is how you will get consistent results with other lights.
First test：LuminTop ODF30
moon： 11
Low：191
Med：682
High：2140
Turbo：4270
ODF30 Now
moon： 4
Low：124
Med：447
High：1423
Turbo：2850
Standard flashlight test light escaping too much?
Light escaping is not a problem by itself, consistency is when the escaping light is the same fraction for all lights.
Exactly my point.
I just tested with my own BLF348 and it read 1 lumen higher with a custom centering ring vs the smallest normal ring.
And what about with your lux meter slightly pulled out?
The meter position will be relative, it will effect all the readings by a fixed percentage. It basically becomes part of the calibration.
That said I do not recommend using the meter position as part of the calibration as it is too hard to keep it consistent over the long term.
Adding a little electrical tape to the inside of the 1st bend would be a better way to fine tune the readings since that will not be effected by bumping the sphere.