Discrepancy in lumens claimed and measurement at flashlightreview.com

[KEK]

I'm contemplating getting either an AE PL24 shorty or AE Xenide 25W for work. So I went looking for reviews and comparisons with other known/used lights in performances and specs. One rather glaring discrepancy is that at flashlightreview, Doug posted that the total output of a surefire 10X as 525 on his chart (advertised 500 lumens) whereas the output of an AE PL24 shorty is 400 (advertised 1000+ lumens).
I don't think that's going to affect my choice but it's just something that I find odd. Can anybody shed some light on this?

 

[2xTrinity]

A lot of manufacturers simply quote "best case" manufacturer numbers for the amount of light coming from the lamp, rather than actually taking a real measurement. About 30% losses in lumens can be expected from reflector/bezel losses, and more losses can be expected as the manufactuer's data sheets are usually based on ideal temperatures/lab conditions that wouldn't be reproduced in a real-world light.

Some vendors lack the instruments to make these measurements themselves, but IMO they should at least say that the lumen figures are based on the bulb specs, and real world figures will likely be lower. Failure to at least state as much is dishonest advertizing.

In the case of companies like Surefire, and Pelican, they not only measure the real lumen output, but their listed numbers a guaranteed minimum -- typical lumens are quite often higher than claimed in their case.

 

[NAW]

Flashlightreviews isn't reffering to lumens when it says output. The only way to measure lumens is to have an IS and Doug (flashlightreviews) does not have one. The device he uses to measure is different.

Hope this helps

 

[Illum]

Doug [flashlightreviews] uses a homemade light meter from a milk carton, his reviews and numbers are based on "direct comparison" only....theres no research backed unit [such as Lumen] in which the numbers are based on, so each light acts as a benchmark to compare other lights to.

when comparing "direct comparison" numbers its best to leave some buffer
and when it comes to reading marketing specs, believe nothing that you read and believe only what you found through experience. marketing has made me distrust every label on any object including the flashlight. I've grown suspicious on whether thats milk I'm drinking or are those really shoes I'm wearing...and what is this "monitor" I'm looking into?!

 

[nerdgineer]

FLR has a section describing the results of trying to relate QB lumens to real lumens. The best guess was that 1 QB output measurement is about 1.62 lumens (if a HID is like a incan light).

 

[Handlobraesing]

A lot of manufacturers simply quote "best case" manufacturer numbers for the amount of light coming from the lamp, rather than actually taking a real measurement. About 30% losses in lumens can be expected from reflector/bezel losses, and more losses can be expected as the manufactuer's data sheets are usually based on ideal temperatures/lab conditions that wouldn't be reproduced in a real-world light.

Some vendors lack the instruments to make these measurements themselves, but IMO they should at least say that the lumen figures are based on the bulb specs, and real world figures will likely be lower. Failure to at least state as much is dishonest advertizing.

In the case of companies like Surefire, and Pelican, they not only measure the real lumen output, but their listed numbers a guaranteed minimum -- typical lumens are quite often higher than claimed in their case.

Best case figure is non-sense. I think 95% should be the cut off.

So, if they tested 10 lights and came up with:
100,105,95,115,108,103,125,85,90 and 92 they should be able to say:
so 95% would be within 79.2 to 124.4,

Then, they could rate it: 101.8 lumens (x bar value) (79.2 lumens minimum for 97.5% of product). Obviously, a randomly selected sample of at least 50 units is desirable to get the x bar as close as you can to µ.

Given the same operating conditions, standard deviation isn't so great on incandescent bulbs, so getting the spec is simple, but with LEDs, I think statistically derived specs are critical.

 

[mtbkndad]

A lot of manufacturers simply quote "best case" manufacturer numbers for the amount of light coming from the lamp, rather than actually taking a real measurement. About 30% losses in lumens can be expected from reflector/bezel losses, and more losses can be expected as the manufactuer's data sheets are usually based on ideal temperatures/lab conditions that wouldn't be reproduced in a real-world light.

Some vendors lack the instruments to make these measurements themselves, but IMO they should at least say that the lumen figures are based on the bulb specs, and real world figures will likely be lower. Failure to at least state as much is dishonest advertizing.

In the case of companies like Surefire, and Pelican, they not only measure the real lumen output, but their listed numbers a guaranteed minimum -- typical lumens are quite often higher than claimed in their case.

Actually most manufacturers don't even try to list best case scenario.
They just list bulb lumens or LED lumens with no consideration to what comes out of the front end of a specific light.

That is why with the Amondotech Illuminator and N30 and L35 I specifically stated bulb lumens are being mentioned in their specs.

With the AI I did a comparison in a large light box I made that can handle spolights.
This comparison was done with one reading of the bulb itself and then another reading of the AI. The AI puts roughly 75% of the bulb lumens out the front end of the light.

The issue is not that manufacturers can't do tests to measure relative light output. Rather they just do not care to. People that steer away from lights from companies like Surefire and Pelican just because the lumen "numbers" are lower then other brands are making
a big mistake.

The important things is to get any given manufacturer to define their terms when making claims about light output.

More lumens in and of itself is not always better anyway .

Take Care,
mtbkndad :wave:

 

[KEK]

Thanks for all the feedback guys. I've done some digging (googling) since the posting and the writeup at the first site from the search on lux meter produced by globalspec (http://optical-components.globalspe...nents/optoelectronics/lux_meters_light_meters) stated that "... The standard color temperature for the calibration of most lux meters is 2856° K, an amount that is more yellow than pure white." At the second site (http://www.action-electronics.com/lightmeter.htm) , one of the model apparently can yield better result up to "±4% accuracy is enhanced by selecting lighting type (Tungsten/Daylight, Fluorescent, Sodium, Mercury) ". I assume that means recalibrating the peak of the reading to different color temperature (and perhaps the width of the spectrum as well).
Doug came up with 2 different coefficients (X1.39 for LEDs and X1.62 for incand) for converting his output measurements to lumen estimation. And he acknowledged that the discrepancy between LED and incandescent measurements may be due to "different spectral detection characteristics of the sensor in the meter". Nerdgineer suggested that the incand coefficient (X1.62) is used for HID. That still puts PL24-S with 20% lower output than 10X :sigh:. I'm a big fan (in spirit that is) of HID lights so I'll just optimistically assume that there is still some uncovered specturm difference from LEDs and incandescents such that the actual coefficient for HID is different from these 2 figures (hopefully >2 :naughty, and that 1300 lumens in HID is actually not worse off than 500 lumens in incandescent lights (that is not to discount or disrespect all the fine technical work done by Doug and many more of you out there).
And of course there's the statistical approach pointed out by Handloraesing and many of you that's probably not used by the manufacturers so there's probably a glaring inconsistency among all claimed outputs, and many many many more.... :hairpull:
Well I guess if I let all these things bother me I'll never buy anything so I'll just stick to my original plan. I would think that 400 lumens or 1300 lumens is a huge leap anyway considering the fact that i'm using an E1L now.
Thanks guys

 

[mdocod]

at FLR... you'll notice he has a disclaimer of sorts that the method is not calibrated to accurately compare LED sources, to incan sources, to IR sources, etc etc... so it could be logically assumed that the equipment is not going to accurately compare HID to incan output spectrum's.. the sensitivity to incan may be greater, causing it to look better on the charts than it is in reality.

 

[NAW]

If its of any help here are some beamshots of the AE 24W VS a Surefire M6. The only problem is, he did not state whether it was the 250 lumen lamp assembly or the 500 lumen lamp assembly used in the M6.

http://www.candlepowerforums.com/vb/showthread.php?t=140922