Texas_Ace BLF Calibrated Lumen tube / Sphere No math skills needed - Several spheres still available

I’ve done the tests textbook like you wished, they get similar readings to ceiling bounce tests. In which the high lumen lights don’t seem to get as close to manufacturers claims, especially floody lights. They do if I lift the light towards the ceiling to match the hotspot of the light I’m getting a figure from. I.e. The DT70 gets 13000 and the X45 gets 11000 by lifting the light to match the spill area I got it to the widely accepted 14500 region.

All I can think is the more light you feed the TA tube the more light it retains, especially with all the diffusers, bounce back, twists and diffuser material, obviously the tube will favour certain beams and lumen ranges.

Though you doubt it it could be my meter playing up though, not being efficient??

KG, I’m just gonna look at one light at a time.

On that X45, turbo is rated at 9,000 lm and turbo max is 16,500 lm.

Are you reading 8,000 lm on turbo max?

8000 lm seems like it might be on the first turbo setting.

If your reading 8000 lm on turbo max, what does it read on the other turbo setting, 4000 lm?

I would expect Acebeam to be a bit closer to spec than most other brands.

Also, as just a general thought, batteries with high internal resistance (like protected cells) or cells like Panny B that can’t do high loads can definitely prevent a light from doing it’s best output.

Like if you had 4 Panasonic NCR18650B 3400mah in your X45 then it would probably not be able to go into the turbo Max mode. Just as an example.

Ceiling bounce tests ONLY work if you take the readings in exactly the same method, in the same room, with the same things in the room. Moving the light higher is skewing the numbers, the light simply reads lower then the rating.

So if the numbers agree when you take the ceiling bounce test from the same place, then that confirms that the sphere is reading correctly.

The reason we use a sphere is to get readings between lights that show the actual numbers, not to confirm the readings that they came with.

Also, ditto Jasons question, I was about to post the same thing. Are you sure the X45 is really on turbo max? That would make a lot more sense if it was not.

Can you please list the cells you are using and if they are fully charged?

X45 cells are the Acebeam ones
DX80 cells is proprietary
D4/TN40vn cells are VTC5A
Q8/MF01/Meteor cells are 30Q’s
DQG Tiny cell is AW
PT16/Timeback II use an 18350 Aspire cell
X5 uses a Nitecore 650mah cell
X80 uses the Acebeam cells
Gen II uses the Acebeam cell

I try to use the best or what they come with. All the batts were fully charged for that list.

The X45/80 needs to be checked by someone as they fire most of their lumens straight to the tubes sidewalls.

I calibrated the TN42 at 2200 lumens using the ceiling bounce phone app, the TN40vn with VTC5A hit over 9000 lumens when compared, on your test it managed 6500, vinh measured it at 8750 if memory serves.

Yes, I turbo maxed the X45 and X80, full batts from cold.

Since I don’t have a calibrate light, I have to use the stock fenix tk15 1000 lumen for educated lumen… so I got 1060 lumen @ 20 seconds… I am fine with it…

Do you have a Noctigon, they should really fire in the 6500-8500 range depending on led.

Take a high output light and measure the difference between a high and low mode on the same light in the tube and using the ceiling bounce method. Is the ratio the same? If it is, the meter is linear.

The tube as a physical object will always be linear with the same light, unless it actually deforms (melts). Although it is always possible for an integrating device to have errors with extremely floody or throwy lights.

Hey kg, don’t compare your lumen with the Vinh number… just focus on the getting the calibration right…

You will not see 8700 lumen ( tn40vn) with the TA tube. Trust me!

When you say you calibrated the TN42 with a phone app, how did you manage this?

The entire point of a calibrated lumen sphere is that it is VERY VERY difficult to calibrate a measuring device without a standard.

Phone lux sensors can vary wildly in the readings as well.

It is also well known that ceiling bounce tests are anything but precise.

It is also difficult to get a reading on the emisar d4. Lumen drop quickly… FYI

’To me’… this seems to be much ado about basically nothing. :person_facepalming:

This is a hundred dollar lumen tube. And while it is pretty darn accurate, it is not a professional sphere.

AND…. if factory quoted lumen specs of particular lights are trying to be ‘verified’, you are wasting your time. ’Most’ specs quoted by light companies are NOT CORRECT & in many cases they outright lie.

So this whole thing has me puzzled as to what the problem is. Maybe I am just to dumb to understand…… maybe not.

Hey kg, i am looking at your lumen number., look like you are taking turn on or otf…

Not on my XP-L HI 6500k Emisars, they stay pretty cool. My ceiling bounce app records turn-on lumens too.

The app displays lumens on the screen and you dial in what the true reading should be, i.e if a “1000” lumen Fenix reads 500 lumens change the app reading to a 1000 lumens!!! If you’re comparing a TN42 to an Acebeam K70 then it should be pretty close.

It’s like comparing the same two Fenix lights, one will be 1000 lumens the other will be 998 etc.

Just never use a flooder on a thrower calibration!!

While it sounds very simple and straight forward, it simply is not that simple.

I have measured hundred and hundreds of lights on a sphere yet my old calibration was still off by ~30% because all the lights read differently and it is impossible to know which one is correct.

So I ended up biasing my readings according to the readings I read on online reviews, which caused the numbers to be even further off.

Also since I don’t think you have read this thread, something most people do no know is that Cree gives a 14% tolerance rating on all of their LED’s we use in flashlights.

That means that is the absolute minimum tolerance any factory light can have when it comes to the accuracy of the lumen rating since they do not measure each light.

Add in a few percent for the driver components and it is easily 20% variance between lights of the exact same model just based on the tolerances of the components.

this is just one of the many reasons why you can not trust rated specs.

So do you have a fenix light? Have you taken any readings with it? Please take readings with any fenix lights you have so we can compare.

While not exact, they are the best factory lights we have found so far. They are at least semi-consistent.

Look like your tube is working properly… i would test it on some stock fenix lights to double check…

I tested my thrunite tn40vn 8700 lumen… i got 7500 @ turn on… 6860 lumen @ 20 seconds…

Just tested the Imalent DT70 at 9200 lumens.

It’s a light with a 16000 lumens claim that I and many other reviewers measured 12000-14000 lumens from. I never heard of a test yielding such a low figure.

Keep in mind that almost all lumen readings you see online are not ANSI and are reading too high. I was one of them.

If you want to read higher numbers, simply add 30% to the numbers and you will have my old calibration which is pretty close to most other online lumen readings you see, but it is not ANSI.

That would put you right at the numbers you expect as well, even if they are not correct.

Basically we need to know if you want real ANSI numbers or if you want numbers that sound good / line up with other people online?

It sounds like you are already doing pretty good for ANSI numbers, if you want to compare your number online, then you need to add about 30% to them.

Should be more @ turn on. Vinh estimates conservatively. For a split second I recorded 10500 with fresh VTC5A.

My X45vn hit 28000 at turn on, vinh claimed 25000, now if you have that!