I haven’t run turbo on my e07 no more than 1 minute… I did see lumen decrease as much as 800 lumen on my emisar d4s…
Do you mean a permanent lumen decrease? Permanently reduced peak output even when the light is cold and turned on in turbo with a fresh cell?
I’ll try some different 30Q’s today inc button top 30Q’s.
Mine also makes a loud click. It is quite offputting. Did anyone figure out what was causing it?
Never figured it out on mine.
However, the sound seems to have gone away on its own. My light no longer makes a clicking sound.
Comparing two calibrated measurements to each other, it’d be reasonable to assume a difference in the cell or the light. However, there is something else here which may account for a lot of the difference. A bunch of people on BLF have attempted to standardize measurements by getting reference lights from maukka for calibration, so their lumen measurements are typically within a few percent of each other most of the time. There have also been some efforts by djozz and Texas Ace and a few others to provide quality integrating devices or teach people how to make them.
But one person in particular rejected all that and did their own thing, insisting everyone else was wrong, and created enough drama to prompt admin intervention. It was pretty unpleasant for everyone involved.
That’s pretty much in the past now, but can still confuse people who aren’t aware of it. Basically, there are multiple definitions of a lumen floating around. There’s the calibrated scale used by maukka and djozz and quite a few other people, there’s the cpfselfbuilt scale which was popular in the past and still lingers, and there’s the KG lumen… which reads higher than everyone else and involves some pretty unusual equipment. So, to compare those to other people on BLF, multiply by about 0.69… ish. That’s the conversion factor people worked out a few months ago when all this was happening.
Going by that, I’d expect about 4230 lm on a 30Q cell.
The 219B option is no longer present on FF website… Am I too late? 
Btw there are now 2 customer reviews on their site.
Added:
Sony VTC5A button top
Feiyutech button top
Samsung 30Q button top
I tried a few different versions of the same battery too with a few small gains. I tried 3 more VTC5A as I thought 7,050 seemed too high compared to the other 18650’s tested and the VTC5A were all around 300 lumens less than the first VTC5A I tried, but the first VTC5A was measured on a hotter day lol
Turn on measurements so far E07 XPL 6500K:
Feiyutech BT - 3,500 lumens
Samsung 30Q BT - 5,030 lumens
Ampsplus Q30 BT – 5,250 lumens
Sony VTC6 FT – 5,300 lumens
sony VTC5A BT - 5,750 lumens
Samsung 30Q FT – 6,130 lumens
Sony VTC5A FT – 6,700 lumens
Samsung 40T – 7,690 lumens
Samsung 30T – 8,300 lumens
We have one more in a loop of five people to test with the same set of Maukka’s calibration lights then I will be sending the calibration lights to a known manufacturer for testing on their professional sphere. The idea is to standardize between our style of lightbox (all made by the same two guys) and typical factory readings. So far our results (4 of us) have been fairly close with the original box they made being closest to Maukka’s own findings.
Of interest to note, the two lights are not off in our boxes by the same margin. So it is still an educated guess where Maukka’s own light sphere would place a light with much higher output than the 290 lumen S2+…. such as the Fireflies E07.
As has already been posted, it is not easy to get duplicate readings on any given light even using the same cell… variables confuse the outputs to the point it’s an assumption by averaging to come up with any number. And so it goes… does the World Record holder in the 100M (or any event) ALWAYS post the same time? Any race car? When you finally get your own light box and start testing you realize just how liquid all the test results really are.
Yesterday I received two identical lights in the mail. I proceeded to make a series of changes in one of them, starting with heavier leads and working up through spring bypasses to an FET swap. I saw output rise to nearly double and then beyond with the application of a better cell. This is really where the lightbox pays its way, assisting in determing what modifications are worth doing. These 2 identical lights were advertised at 100,000 lumens. The stock light is showing 1320 lumens on my box, the one I’ve been tweaking is up to 5600 so far… this still using the XM-L T6 emitters. Determining mod success, not stone-carving rights… that’s really the true measure of the home made light box. 
The E07 is the first light I’ve had (from over 100) that has an output that varies so dramatically with each cell. I mean a 5000 lumens between 2 different cells is unbelievable.
Even the same cell is getting different results daily, today my 40T can’t get past 7,690 lol
Maukka’s calibrated lights prove nothing other than modifying a sphere/box/tube to only measure 1 flashlight/beam profile accurately, it’s absolute folly. I’m sick of hearing about them.
Never heard of Feiyutech, not surprising it performed so poorly. Given the disparity of your tested cells it really isn’t shocking at all how far apart their numbers are. I’ve tested over 5 times the number of lights you speak of and still get surprised sometimes….
With direct drive and 7 emitters, the E07 will exploit the cell and show out any and all weakness. The greater a lights potential the more important proper cell choice.
Is there any chance that Fireflies will correct the flaws of the E07 in the near future? 1-2mm longer tube and correctly positioned screws. No broken optics legs. I really wanted to buy this light but not want to start with full disassembling and correcting problems DIY. If they will be corrected I will buy one but If not I will buy another D4S.
Yes I know it can be solved DIY but I don’t want to spend a light 70-80$ that I need to do DIY repairs instantly out of box to use it. I know this solves the short tube problem but it makes a gap between head and tube. And if I know it is there I can’t get my eyes off that gap 
The tube is long enough for a real 40T and MANY other batteries. Its the fake 40T Lishen they sell that doesnt fit
It seem the complaints have been heard and they are trying to fix them, a good news! like you i'm waiting a little bit for a better finished product.
Do you usually notice a cell gets its peak output when used for the first ever time?
I got 8300 on a brand-new never used 40T but on every test since it’s been around 7700
Yes, the cell settles in with some use as do the LED’s themselves. Enough hard use and the cells no longer charge to the same Voltage as when new.
And then there are those times when one particular brand gives the most amperage but not the most lumens… yes once in while a brand or particular model will give the most lumens, sometimes by quite a spread, but show less tail amps being pulled.
By chance do you think it could be the contact from the battery? Sometimes mine appears to have a weak turbo, then I break connection on the tail cap and tighten it down and try again with huge difference in output.
Clean tube end contacts with a clean wipe. It can make surprising jumps in lumens sometimes.