degarb
Flashlight Enthusiast
Calculating Lumens Via Relative Diffused Lux : My Dlux Method
Hypothesis: If you find a perfect diffuser (loss/absorption by the diffuser is immaterial), any 50 lumen light will have same lux of "x" at a fixed distance, as would any 100 lumen light have double that "x" lux,etc.
Look, Ma, no Lumen Meter! This means that if you measure the "dlux"- diffused lux- of a trusted light manufacturer with a known luminous output and flat regulation, then, measure the dlux -diffused lux- of the unknown light, the dlux ratio multiplied against the known luminous output, should yield a lumen number be within a few percent of the real lumen output. The purpose: I read lumen meters are expensive while a lux meter can be had for under thirty dollars US.
Limitations on Accuracy, My List:The "Flat Current Driver of Known Light" quality tolerance,variation within each bin of the known LED, dishonest trusted light makers, and differing led color v. uneven absorption of certain wavelengths by your white diffusing medium. Notwithstanding these limits, my bet and initial tests show about a 7 percent variation from expected lumen values-despite my hasty, initial, sloppy experimental conditions.
Experiment tools: lux meter,sheet of white ink jet paper for diffuser, milk crate to fix the distance from light to meter, chicken wire to stiffen paper, cardboard to steady lux meter, pen and pad to write numbers down.
Preliminary Notes: The distance to light meter does not matter, as long as exactly consistent. I recommend half to a third a meterish Loss through paper does not matter, but should probably be white, in hopes of equally blocking/passing all of the color spectrum. Milk crate on side may make number reading easier, flashlight head needs to touch paper and shine down directly onto diffuser, where angle is exactly level. An empty fish tank or a white tube would work better, perhaps. Fenix is my trusted manufacturer, known for efficient, flat, proprietary drivers.
Experiment procedure: Set milk crate on side, put lux meter on cardboard on bottom in exact place,put chicken wire on top part of crate, and then paper on top of chicken wire, finally each light head will need to be pressed gently onto paper for measuring the dlux number. First , measure and write down the dlux of the known light, then do same with unknown light.
The Math: Divide the dlux of the unknown light by the known light, this is the dlux ratio. Multiply the dlux ratio and the known lumen out put and you have an approximate lumen value of the unknown light.
For Example: If the dlux ratio is 2 with a known 50 lumen light, the unknown light is 100 lumens;else, if the dlux ratio is .5, the unknown light is 25 lumens.
Alternative setup: Take a 1.5 foot card board tube, glue paper on one end, white panty hose on the meter end, paint inside white. Use tube to set the distance, uselevel to make sure angle stays consistent.
Commentary: If a flashlight reviewer verifies the dlux ratio lumens - over just trusting the manufacturer's package - , then the reviewer tells us the lux in hotspot (radius at 1 meter) and spill--and tailcaps current, and temperature...Then, and only then, does the subjectivity go away in the review.
So far, with lack of privacy, I have not the time/opportunity to do the above experiment precisely myself,just approximations using knee highish level. But I am pleased with calculations within 7 percent of expected values (known trusted light to trusted light of same manufacturer. I have been shocked at how low the dlux lumen calculated output of several "brick and mortar" direct drive; and have pretty much verified things I suspected with my eyeball meter. For example I have an xpg r3 and and xpc, both rated at 110 lumens by their manufacturer, both 87 calculated dlux lumens (against a Fenix Light), both wildly differing beam patterns and throw. Apparently, this verifies that brick and mortar manufactures indeed do not account for the 20% to 30% optical loss,while Fenix does.
I never trusted the milk jug or ceiling bounce methods, as they fail to account for angle of light bounce v. beam pattern, which should throw off numbers. Also more importantly, they fail to approximate any lumen number.
Also, with my method, the calculated lumens via dlux ratio should ideally be averaged among several known lights of flat regulation (say, 5), to get past the unknown limitation factors I mention above.
If we, as a community, find there is too much driver variation by all manufacturers, inaccurate lumen numbers by all manufactures, then we could say there is no "known" lights—who can tell which lights can be trusted. Still, perhaps, we could standardize a dlux value--.5 meter with inkjet paper of a certain brand/weight. Hopefully, this is not the case.
Enjoy calculating.
Hypothesis: If you find a perfect diffuser (loss/absorption by the diffuser is immaterial), any 50 lumen light will have same lux of "x" at a fixed distance, as would any 100 lumen light have double that "x" lux,etc.
Look, Ma, no Lumen Meter! This means that if you measure the "dlux"- diffused lux- of a trusted light manufacturer with a known luminous output and flat regulation, then, measure the dlux -diffused lux- of the unknown light, the dlux ratio multiplied against the known luminous output, should yield a lumen number be within a few percent of the real lumen output. The purpose: I read lumen meters are expensive while a lux meter can be had for under thirty dollars US.
Limitations on Accuracy, My List:The "Flat Current Driver of Known Light" quality tolerance,variation within each bin of the known LED, dishonest trusted light makers, and differing led color v. uneven absorption of certain wavelengths by your white diffusing medium. Notwithstanding these limits, my bet and initial tests show about a 7 percent variation from expected lumen values-despite my hasty, initial, sloppy experimental conditions.
Experiment tools: lux meter,sheet of white ink jet paper for diffuser, milk crate to fix the distance from light to meter, chicken wire to stiffen paper, cardboard to steady lux meter, pen and pad to write numbers down.
Preliminary Notes: The distance to light meter does not matter, as long as exactly consistent. I recommend half to a third a meterish Loss through paper does not matter, but should probably be white, in hopes of equally blocking/passing all of the color spectrum. Milk crate on side may make number reading easier, flashlight head needs to touch paper and shine down directly onto diffuser, where angle is exactly level. An empty fish tank or a white tube would work better, perhaps. Fenix is my trusted manufacturer, known for efficient, flat, proprietary drivers.
Experiment procedure: Set milk crate on side, put lux meter on cardboard on bottom in exact place,put chicken wire on top part of crate, and then paper on top of chicken wire, finally each light head will need to be pressed gently onto paper for measuring the dlux number. First , measure and write down the dlux of the known light, then do same with unknown light.
The Math: Divide the dlux of the unknown light by the known light, this is the dlux ratio. Multiply the dlux ratio and the known lumen out put and you have an approximate lumen value of the unknown light.
For Example: If the dlux ratio is 2 with a known 50 lumen light, the unknown light is 100 lumens;else, if the dlux ratio is .5, the unknown light is 25 lumens.
Alternative setup: Take a 1.5 foot card board tube, glue paper on one end, white panty hose on the meter end, paint inside white. Use tube to set the distance, uselevel to make sure angle stays consistent.
Commentary: If a flashlight reviewer verifies the dlux ratio lumens - over just trusting the manufacturer's package - , then the reviewer tells us the lux in hotspot (radius at 1 meter) and spill--and tailcaps current, and temperature...Then, and only then, does the subjectivity go away in the review.
So far, with lack of privacy, I have not the time/opportunity to do the above experiment precisely myself,just approximations using knee highish level. But I am pleased with calculations within 7 percent of expected values (known trusted light to trusted light of same manufacturer. I have been shocked at how low the dlux lumen calculated output of several "brick and mortar" direct drive; and have pretty much verified things I suspected with my eyeball meter. For example I have an xpg r3 and and xpc, both rated at 110 lumens by their manufacturer, both 87 calculated dlux lumens (against a Fenix Light), both wildly differing beam patterns and throw. Apparently, this verifies that brick and mortar manufactures indeed do not account for the 20% to 30% optical loss,while Fenix does.
I never trusted the milk jug or ceiling bounce methods, as they fail to account for angle of light bounce v. beam pattern, which should throw off numbers. Also more importantly, they fail to approximate any lumen number.
Also, with my method, the calculated lumens via dlux ratio should ideally be averaged among several known lights of flat regulation (say, 5), to get past the unknown limitation factors I mention above.
If we, as a community, find there is too much driver variation by all manufacturers, inaccurate lumen numbers by all manufactures, then we could say there is no "known" lights—who can tell which lights can be trusted. Still, perhaps, we could standardize a dlux value--.5 meter with inkjet paper of a certain brand/weight. Hopefully, this is not the case.
Enjoy calculating.
Last edited: