Cheap way of measuring lumens

I want to start measuring the lumens of my flashlights.
What is the best way to go.
It has to be reasonable in price (is 100$ enough?)
and acceptable accurate (give or take 5%?).

Unfortunately, luminous flux (lumens) has to be measured with an integrating sphere, which is not cheap.

Fortunately, more and more flashlights today are advertised with realistic lumen ratings. (Back in the 2001-2006 time period, it was common for flashlights to be "rated" as 2-3x as bright as they really were, typically due to numbers being taken from an emitter spec sheet not accounting for optic losses, underdriving, etc.)

I've always felt you could probably devise a system using a camera with manual settings. Lots of trial and error for sure, and what to use for your standards. I'd probably start with measuring ceiling bounce differences.

Geoff

Flying Turtle said:

I've always felt you could probably devise a system using a camera with manual settings. Lots of trial and error for sure, and what to use for your standards. I'd probably start with measuring ceiling bounce differences.

Geoff

Click to expand...

I think ceiling bounce shots are an ideal way to get a sense of total output and relative comparison light to light. Here are a few I did a while back.

RichS's method (as shown in that excellent thread he has linked) is a very effective way to show up the differences between one light and another. Without an integrating sphere, a comparison ceiling bounce shot vs. another light that has a known output is the most reliable alternative.

It you want relative differances, I know some of the Canon camera's can be "hacked" so you can see the lux readings...or you could also build a cheap "dark box" and download the "Lux Meter Pro" app for an IPod Touch/IPhone. I know it's not that scientific, but it will show you relative differences in brightness.

Build Pictures of My 16" and 24" Homemade Integrating Spheres
http://www.candlepowerforums.com/vb...quot-and-24-quot-Homemade-Integrating-Spheres

Just build my own "poor man's" integrating sphere
http://www.candlepowerforums.com/vb...y-own-quot-poor-man-s-quot-integrating-sphere

I'm sure you could use teh lumisphere on a sekonic lightmeter in some fashion.

you'll get readings in F-stops and 10ths of power ratings though.

I downloaded "Light Meter" for my Samsung Galaxy S2. It works.
This program allowed me to calibrate the light myself.

I took a TK35 and placed it on a tripod at 1 meter from the sensor in the Samsung. Then calibrated it till it gave me 12 lux at 12 lumens (low mode) and 820 lux at 820 lumens (Turbo mode).
Then I tried it on the Sunwayman T20C and it gave a reading at 430 lux in the highest mode (438 lumens)
Not scientific, but fun to do
There should be a more easy way to read acurate Lumens

RichS,
Thanks for the link to your ceiling bounce shots, it is a lot of work, but payed of in the end. Probably the cheapest and most honest way to compare lights for guys like us.

I use an excellent Minolta Lightmeter and for my standard, a Quark Turbo which is 230Lu. in a ceiling bounce test. This is primitive, to be sure, but it tells me little things i.e. my recently acquired HDS 200T. reads slightly higher than the Quark and slightly less than a Dereelight CL1H V4 which is advertised as 240Lu. :thumbsup:

I just bought this:
http://www.cross-mark.com/digital-light-meter-ct1330b-p-1669.html

Can't go wrong for this price. Just to see how it works and play around with some ceiling bounce tests. :thumbsup:

Grizzlyb said:

I downloaded "Light Meter" for my Samsung Galaxy S2. It works.
This program allowed me to calibrate the light myself.

I took a TK35 and placed it on a tripod at 1 meter from the sensor in the Samsung. Then calibrated it till it gave me 12 lux at 12 lumens (low mode) and 820 lux at 820 lumens (Turbo mode).
Then I tried it on the Sunwayman T20C and it gave a reading at 430 lux in the highest mode (438 lumens)
Not scientific, but fun to do
There should be a more easy way to read acurate Lumens

RichS,
Thanks for the link to your ceiling bounce shots, it is a lot of work, but payed of in the end. Probably the cheapest and most honest way to compare lights for guys like us.

Click to expand...

Hi GrizzlyB

Where did you get this app? The ones I have downloaded from google app store don't work properly on my Galaxy S2 - each app only gives me one of 3 lux values depending on the light intensity - either 10, 100 or 1000 lux. By contrast my son's Huawei Blaze (U8510) seems to work perfectly giving accurate-ish results down to 0.1 lux.

Thanks Chris.

Measuring Lumens

I know that measuring lumens can be tricky, but does anyone have a quick and dirty way that at least puts you in the ballpark?

Re: Measuring Lumens

EUREKA! I HAVE DONE IT! A 'BALLPARK' LUMEN MEASURING METHOD! I will smugly and egotistically claim ALL CREDIT for this method, as I have found it CONSISTENT with all my lights without further ado, here you go! (insert complementary ego stroke here)

Get some diffusers for lights of different sizes. 22mm, 26 mm, 50 mm, etc. Make sure they're quality plastic ones. I used a Kirkland water bottle at the white stripe for centering, PM if you're curious how i did it. I will be making a video soon!

I noticed that if I placed ANY of the diffusers 90 degrees horizontally centered at about 4 and 1/4 inches above the table AND put a cheap LX1010B LUX METER on the table directly under the diffuser, I could put corresponding sized flashlights into them and get a number that reads ballpark lumens. I tested this with 22mm diffusers with SWM v11r, Jetbeam PC10, Jetbeam RRT0xml, Xeno E03, and 50mm diffusers Nitecore TM11, TK41, Defiant Flashlights.

Settings on the lux meter MUST be set for x10 lux (or 20000 on the LX1010B) for anything under 1000 lumens, and X100 lux (or 50000 on the LX1010B) for anything under 10000 lumens. The number you see are the lumens you get at the x10 lux setting, and the number you see is lumenx10 at the x100 lux setting. If you see 200 at the x10(20000) reading, it should be understood as 200 lumens. If you see 200 at the x100(50000) reading, it should be understood as 2000 lumens. I have yet to build a good standing rig to make sure all diffusers remain level and such, but if you wanted a quick and dirty way to do it, this is IT.

Here are some 'lux lumens' I get from my method:
TM11:
Low: 205 david luxlumens
Medium: 535 david luxlumens
High: 1119 david luxlumens
TURBO: a whopping 2190 lumens! (fully charged cells of course)

TK41:
Low: 14 david luxlumens
medium: 110 david luxlumens
high: 325 david luxlumens
Turbo: 765 david luxlumens

Sunwayman v11r IMR:
Turbo: 482 david luxlumens

RRT0 IMR:
Turbo: 525 david luxlumens

Jetbeam PC10 IMR:
Turbo: 487 david luxlumens

Klarus ST20:
Moonlight: 2 david luxlumens
Low: 13 david luxlumens
Medium: 123 david luxlumens
High: 240 david luxlumens

Xeno e03 NW xml AA:
Low: 6 david luxlumens
medium: 40 david luxlumens
high: 94 david luxlumens

Thrunite TiS aaa
Single: 66 david luxlumens

Defiant 5c 650 lumens:
semi drained cells: 573 david luxlumens

IT WORKS! Please go test it for your self. I have finally contributed something awesome for the layman at CPF!

Re: Measuring Lumens

dc38 said:

EUREKA! I HAVE DONE IT! A 'BALLPARK' LUMEN MEASURING METHOD! I will smugly and egotistically claim ALL CREDIT for this method, as I have found it CONSISTENT with all my lights without further ado, here you go! (insert complementary ego stroke here)

Get some diffusers for lights of different sizes. 22mm, 26 mm, 50 mm, etc. Make sure they're quality plastic ones. I used a Kirkland water bottle at the white stripe for centering, PM if you're curious how i did it. I will be making a video soon!

I noticed that if I placed ANY of the diffusers 90 degrees horizontally centered at about 4 and 1/4 inches above the table AND put a cheap LX1010B LUX METER on the table directly under the diffuser, I could put corresponding sized flashlights into them and get a number that reads ballpark lumens. I tested this with 22mm diffusers with SWM v11r, Jetbeam PC10, Jetbeam RRT0xml, Xeno E03, and 50mm diffusers Nitecore TM11, TK41, Defiant Flashlights.

Settings on the lux meter MUST be set for x10 lux (or 20000 on the LX1010B) for anything under 1000 lumens, and X100 lux (or 50000 on the LX1010B) for anything under 10000 lumens. The number you see are the lumens you get at the x10 lux setting, and the number you see is lumenx10 at the x100 lux setting. If you see 200 at the x10(20000) reading, it should be understood as 200 lumens. If you see 200 at the x100(50000) reading, it should be understood as 2000 lumens. I have yet to build a good standing rig to make sure all diffusers remain level and such, but if you wanted a quick and dirty way to do it, this is IT.

Here are some 'lux lumens' I get from my method:
TM11:
Low: 205 david luxlumens
Medium: 535 david luxlumens
High: 1119 david luxlumens
TURBO: a whopping 2190 lumens! (fully charged cells of course)

TK41:
Low: 14 david luxlumens
medium: 110 david luxlumens
high: 325 david luxlumens
Turbo: 765 david luxlumens

Sunwayman v11r IMR:
Turbo: 482 david luxlumens

RRT0 IMR:
Turbo: 525 david luxlumens

Jetbeam PC10 IMR:
Turbo: 487 david luxlumens

Klarus ST20:
Moonlight: 2 david luxlumens
Low: 13 david luxlumens
Medium: 123 david luxlumens
High: 240 david luxlumens

Xeno e03 NW xml AA:
Low: 6 david luxlumens
medium: 40 david luxlumens
high: 94 david luxlumens

Thrunite TiS aaa
Single: 66 david luxlumens

Defiant 5c 650 lumens:
semi drained cells: 573 david luxlumens

IT WORKS! Please go test it for your self. I have finally contributed something awesome for the layman at CPF!

Click to expand...

I think this is akin to the ceiling bounce method in principle, but scaled down to concentrate the light on a smaller more proximal target. The weakness of the ceiling bounce has always been the tendency to over represent hot spots, as more light is reflected straight down at the meters by the central part of the beam on the ceiling, and to underestimate the corona and spill portion of the output at more peripheral areas that are proportionally less likely to reflect back at the meter.

The other weakness for all home made/DIYS devices is the calibration of the meter itself. If the meter is accurate to +/- 10% for example, your lumen measurement cannot be more accurate than that either, as its a limiting factor. So a reading of 1,000 Lux could have been ~ 1,100 or 900 lux for example.

Of course, meters have scales, and, are more accurate in some parts of the scale than others. Typically, mid-scale readings tend to be more accurate, with the accuracy falling at the upper and lower ends of the involved ranges.



If you don't trust the ANSI #'s, and I have a sneaking suspicion that some are NOT done with NIST certified IS equipment, etc....then its a way to at least get a feel for how close they might be...and, its a way to benchmark modded lights, etc.


The issues that could complicate matters would be:

The distances, as the lux drops according to the inverse square law, so a small difference in distance to the meter sensor will have a large impact on the calculated lumens. So the distances would require careful calibration to be applicable as apples and apples for different test runs, etc.



I'm looking forward to your video!!

Re: Measuring Lumens

TEEJ said:

I think this is akin to the ceiling bounce method in principle, but scaled down to concentrate the light on a smaller more proximal target. The weakness of the ceiling bounce has always been the tendency to over represent hot spots, as more light is reflected straight down at the meters by the central part of the beam on the ceiling, and to underestimate the corona and spill portion of the output at more peripheral areas that are proportionally less likely to reflect back at the meter.

The other weakness for all home made/DIYS devices is the calibration of the meter itself. If the meter is accurate to +/- 10% for example, your lumen measurement cannot be more accurate than that either, as its a limiting factor. So a reading of 1,000 Lux could have been ~ 1,100 or 900 lux for example.

Of course, meters have scales, and, are more accurate in some parts of the scale than others. Typically, mid-scale readings tend to be more accurate, with the accuracy falling at the upper and lower ends of the involved ranges.



If you don't trust the ANSI #'s, and I have a sneaking suspicion that some are NOT done with NIST certified IS equipment, etc....then its a way to at least get a feel for how close they might be...and, its a way to benchmark modded lights, etc.


The issues that could complicate matters would be:

The distances, as the lux drops according to the inverse square law, so a small difference in distance to the meter sensor will have a large impact on the calculated lumens. So the distances would require careful calibration to be applicable as apples and apples for different test runs, etc.



I'm looking forward to your video!!

Click to expand...

I agree that my method is in NO way scientific, but merely coincidental. However, the numbers that come up seem to match the manufacturers' 'ansi' ratings...upon 15-30 seconds of initial activation, anyways. I want to do a video with a stand, but have not yet secured a stand on which to mount everything uniformly. I'm thinking of using a chemist's test tube holder stand thingy with different sized clamps

Re: Measuring Lumens

I am definitely waiting to see the video on this

Re: Measuring Lumens

dc38 said:

EUREKA! I HAVE DONE IT! A 'BALLPARK' LUMEN MEASURING METHOD!

IT WORKS! Please go test it for your self. I have finally contributed something awesome for the layman at CPF!

Click to expand...

Good for you.... whatever diffusion method you use, you quickly get a sense of its accuracy by testing a variety of your own lights against the manufacturer spec - and IMHO, your own results will be far more accurate and consistent than what I've seen claimed by certain manufacturer's "ANSI FL1" specs. Those rules have holes wide enough to drive a truck through, not to mention the conservatism the humble manufacturers employ.

I personally use a horizontal bounce (between two walls in a narrow hallway) for diffused sampling, but have also tried the white plumbing pipe method with the same results (piping the light through two 90-degree elbow bends).

Don't forget to reconcile with Selfbuilt's results where you can, although I use a significantly more conservative lumen scale than he.

Do you have any Zebralight AAs you can test?

Re: Measuring Lumens

reppans said:

Good for you.... whatever diffusion method you use, you quickly get a sense of its accuracy by testing a variety of your own lights against the manufacturer spec - and IMHO, your own results will be far more accurate and consistent than what I've seen claimed by certain manufacturer's "ANSI FL1" specs. Those rules have holes wide enough to drive a truck through, not to mention the conservatism the humble manufacturers employ.

I personally use a horizontal bounce (between two walls in a narrow hallway) for diffused sampling, but have also tried the white plumbing pipe method with the same results (piping the light through two 90-degree elbow bends).

Don't forget to reconcile with Selfbuilt's results where you can, although I use a significantly more conservative lumen scale than he.

Do you have any Zebralight AAs you can test?

Click to expand...

Unfortunately, I do not...I have yet to acquire any ZL's. I am looking to pick up an SC52 one way or another, lol. As for my soon-coming video, I will be able to make it as soon as I find my darned camera...Went out during the week and dunno where I might've placed it...

Re: Measuring Lumens

So it has been almost a year since I promised a video, and I still don't have one yet . Found my camera, charging up the battery now. Hopefully will have a video up by sunday, the latest. BTW, Just going to post about a few more lights I've gotten over the course of this year:

MM15vn: Initial activation/15 seconds
MAX: 7540 davidlumens

K40vn: Initial activation/15 seconds
MAX: 1620 davidlumens

Lenser P3vn: .17A at the tail
MAX: 12 davidlumens

Balder SE1: Inital activation/eneloop
MAX: 72 davidlumens

47 Atom: Initial activation/included cr123a
MAX: 152 davidlumens

Jetbeam RRT-01: Initial activation 18650/15 seconds
MAX: 720 davidlumens

**A quick 'clarification' of sorts as to how I concieved the method; since an integrating sphere collects ALL the emitted light in order to measure an even "cross section" of light, I figured that there must've been a minimal distance that I could set a lux meter away from a diffuser that would essentially do the same thing as Quickbeam's/Selfbuilt's mini lightbox setup. Since a diffuser pretty much filters and evens out the beam, there should be a coincedental distance where the readout would give me the approximate representative number of lumens. The new poland spring bottles work fine as a reference point as well, the center of the top of the diffuser placed horizontally should align perpendicular to the the bottom of the label on the bottle. Both the bottle and the lux meter measurer should be resting on the same surface. Again, for anything known to be 1 lumen up to 999 lumens, it should be set to the X10 setting. From 1000 lumens to 10000 lumens set to the x100 setting. Anything less than a lumen, x1 setting. Turn out the lights and give it a whirl!