Some results of measuring with luxmeter

[Swedpat]

With my luxmeter I have measured several of my flashlights and have some interesting? results to provide.

My main intention has been to compare the difference of brightness between different lights by shining up on the roof. Then I get the percentual difference of total light output between the lights. I don't know the actual lumen rating anyway, but it's still interesting, I think.

Nitecore D10 is rated at max 130 lumens. The max brightness is 86% of Fenix L1D at turbomode. IF the true output of L1D is 120 lumens, the true output of D10 is 103 lumens. Not bad at all, but far below the stated output.

Tiablo A9 is rated at 250+lumens, and Fenix TK10 at 225. My estimation with my eyes has earlier been that the TK10 is at least as bright, of not brighter. That's correct. The TK10 is ca 15% brighter than A9.
If the true putput of the TK10 is 225 lumens, it means that the true output of the A9 is 195 lumens. The opposite case would give around 290 lumens for the TK10, which I think we can eliminate.

When I measured the output of my 4E20 Link I found that the brightness between the separate lights varied. The brightest example of E20 is 22% brighter than the dimmest. The total output of the 4 parts is 87% higher than the TK10. If my example of the TK10 has 225 lumens that does mean my 4E20 has 421 true lumens. The day before the TK10 was slightly dimmer and the 4E20 was 97% brighter, however...(not changed the batteries)

My ordinary E20 is 49% of the brightness of the TK10 (not day before). Actually it would give 110 lumens. In comparison to the average value of the different parts of my 4E20 it has 47,6% of the output of the TK10 (107 lumens).

BUT; the average output of all my 5 parts of E20 is 19,2% higher than the L1D at turbomode. Supposing the L1D has true 120 lumens the average E20 has 143 lumens...

Fenix TK20 has 45% higher output as the average E20, which quite well corresponds to the stated 150 lumens. Supposing the average E20 has 109 lumens it gives 158 lumens.

The TK10 at low mode is only 20,6% of Turbomode. If turbomode has 225 lumens the low mode has 46 lumens, not 60.

Lowmode of TK20 is 26% of the turbomode; which will give 39 lumens if the turbomode has 150.
Lowmode of the TK20 is 88% of the lowmode of TK10.
Turbomode of TK20 is nearly 70% of the turbomode of TK10.
The measured difference of turbomode brightness between TK10 and TK20 pretty well correspond to the stated.

The 4E20 is very exactly as bright as the TK10 and A9 together.

The question is if the difference of brightness between different examples of TK10 (and other lights) also is as large as the difference I can measure of all my E20s?
Though I actually expected a slight difference between my E20s I didn't expect such a huge difference.

I love the 4E20, however, it's the brightest light I have ever had. If all 4 parts had the same output at the brightnest, the total lumen rating had been 470 lumen, providing the true output of my TK10 is 225 lumen. I think I would hardly notice that difference, however. But the 87% higher value is significant in comparison to the TK10.

In all these tests I have used fresh (or very short time used) or fully charged batteries.

Complement: recently I tried the Fenix LOD Q5.
Claimed brightnesses: 25, 9 and 75 lumens.

25 lumens level: half of the L1D at the 53 lumens level.
9 lumens level: 75% of the L1D at 12 lumens level.
75 lumens level: 64% of the L1D at 120 lumens level, but 160% of the 53 lumens level.

These numbers corresponds quite well between the LOD and L1D.

Regards, Patric

 

[Marduke]

Lux readings cannot be related to total light output. Lux and lumens are entirely different measurements, defendant upon the geometry of the optics.

 

[Guy's Dropper]

What are lux?

 

[csshih]

there we go:http://en.wikipedia.org/wiki/Lux
a simple wikisearch.

 

[DavidD]

My main intention has been to compare the difference of brightness between different lights by shining up on the roof.

Marduke, you are right in saying that Lux & Lumen are different. However, IF Swedpat means 'ceiling bounce' by 'shining up on the roof', then it can give comparative results. Remember English is not his first/native language.

Without an IS, the only ways we have to measure total brightness are comparative luxmeter numbers in a homemade milk carton contraption and comparative luxmeter numbers while doing a ceiling bounce in a controlled environment.

But then again, maybe Swedpat is taking the lights outside and truly 'shining up on the roof'.?.

 

[Bullzeyebill]

Swedpat, would you please explain your light bounce procedure? Please describe the enviornment where the testing is taking place. How far are the lights from the lightmeter, etc?

Bill

 

[MrGman]

I read this earlier and didn't want to be the first to comment, plus I was on my way out the door. I took his statement to be a ceiling bounce test but still would not consider what he did as accurate to make comparative statements about lumens from various flashlights.

My first guess is he is actually reading the "Lux" output readings from the ceiling bounce directly off the meter.
That cannot be translated to lumens accurately at all.

It can be used simply to compare difference in lux output from flashlight to flashlight when done in the same test. I don't believe it can be used to help establish what the vendor's say their lumens output are, or use one vendor's stated lumens to say another one is X percentage higher or lower (in lumens), just that the lux reading in a ceiling bounce test is X percentage higher and lower and nothing more.

We know nothing about the beam patterns and ceiling textures and how that would affect the readings to try and make comparative lumens determinations.

I won't even get into the lack of calibration issues with such a test.

So to be constructive, I would say Swedpat. Just put a table together that shows the actual lux readouts at the exact same total ceiling bounce distance for the various lights and then put the hot spot readings at 1 meter next to that and simply publish the table for what it is and not try to determine that vendor's lumens readings are high, low or correct.

I am after this Thanksgiving holiday break going to work with Wbp to see about helping to calibrate his home made integration sphere that he uses with a $9K light meter and see if we can get some good readings out of that. He has a bunch of lights. I will also be measuring some of his lights at work to add to my list of real IS lumens readings to help round out the list.

I make no apologies when I say that I don't find ceiling bounce or light "box" readings useable for even comparative lumens determinations.

 

[Swedpat]

Thanks all for your replies,

Yes it's right, english isn't my native language and I am aware I may sometimes fail to express myself correct or perfectly understandable.

I know the difference between lux and lumens. I am using a luxmeter, and have no idea whether it's possible to also measure lumens with it. But I know that if I shine towards an even coloured surface (in this case white painted roof) and measure the reflected light I can read the difference of measured lux and then know the percentual difference of lumens.

My tests are done with the luxmeter placed on my table and holding the lights about 1 meter above the table shining straight upwards to the roof above the sensor. It will not give 100% the correct value (a very slight move will cause a slight change of the value).
But I try to place the beam as close as possible equal position every time. Also I take in consider the width of the beam and try to hold the light at that distance that the beam diameter on the roof will be similar.
I think this method will give at least 90% fair and comparative result.

Regards, Patric

 

[Marduke]

But I try to place the beam as close as possible equal position every time. Also I take in consider the width of the beam and try to hold the light at that distance that the beam diameter on the roof will be similar.
I think this method will give at least 90% fair and comparative result.

Regards, Patric

That will greatly skew the results.

You have also not taken into account the differences between OTF and emitter lumens.

 

[Bullzeyebill]

Swedspat, do not dispair. Keep up your bounce with lightmeter work. Many of us on CPF have been using bounce with lightmeter and lightbox readings to get approximate lumen readings, particularly when we are comparing readings to our own lights, and lights that have tested IS lumen numbers. Quickbeam (flashlightreviews.com) used his lightbox to get some fairly accurate lumen estimates, and when we did the Lightmeter Benchmark passaround several years ago, and later the passaround lights were tested in an IS by LSI labs, he had to make only a small adjustment in his formula numbers to be spot on. Keep up the work.:thumbsup:

Bill

 

[Swedpat]

That will greatly skew the results.

You have also not taken into account the differences between OTF and emitter lumens.

Marduke,

I cannot see how it would skew my results by taking in consider the width of the beam. I understand it would if I didn't do it! A wider projected beam means that the lux readings will decrease because the outer parts come farer from the sensor.

Surely, the most fair result would likely be to put each light into a translucent globe and then place the sensor at such a distance from the globe.

I don't know what does it mean OTF lumens.

Regards, Patric

 

[Marduke]

By changing the position you entering another variable into your "relative" comparison, making it no longer comparative. The ceiling bounce is already dubious when comparing lights of very different beam patterns. An actual light box will give much more consistent relative brightness figures.


OTF lumens and Emitter lumens are the difference between the light produced by the emitter, and the light which actually exits the light. Different manufacturers rate differently.

 

[Flying Turtle]

I agree with Bill. Keep up the good work, Swedpat. I do a lot of ceiling bounce comparing and have considered some arrangement with my camera for measurement. Your setup sounds better.

Geoff

 

[Swedpat]

By changing the position you entering another variable into your "relative" comparison, making it no longer comparative. The ceiling bounce is already dubious when comparing lights of very different beam patterns. An actual light box will give much more consistent relative brightness figures.


OTF lumens and Emitter lumens are the difference between the light produced by the emitter, and the light which actually exits the light. Different manufacturers rate differently.

Marduke,

Thanks for answer. My intention with the tests is to measure the difference of true output between different lights, and therefore I actually recieve a quite good result of the true difference between different lights OTF lumens. Though I cannot know the lumens. For example: if I put on 4 pieces of E20 the total brightness is measured to be VERY exactly the sum of the brightness of them individually. The result isn't always 100% but it's still comparative. I can for example see that the TK10 is close to exactly twice as bright as E20, which well corresponds to the stated number. Also my measurements well correspond to the estimation I have done with my eyes.

Regards, Patric

 

[jenskh]

By changing the position you entering another variable into your "relative" comparison, making it no longer comparative. The ceiling bounce is already dubious when comparing lights of very different beam patterns. An actual light box will give much more consistent relative brightness figures.

I think he is making the comparison better by adjusting the distance to get about the same beam width. What he is after, is the number of lumens reflected from the roof. The number of lumens hitting the roof, is very close to independent of this distance. By adjusting the distance, he is getting more comparable beams.

As I have presented elsewhere, I calculate the lumens based on measuring the lux profile. I get the same lumens independent on the distance from the light and to my measuring plane.
Jens