At the emitter, out the front. What's the deal?

Highbeam101 · Jun 3, 2008

Iv'e been reading the specs on a lot of lights. They list their lumen rating as at the emitter, out the front, or they don't say. I guess there's no way to tell unless you know how to measure the light source. My question: If you have light states 100 lumen at the emitter, is there a way to figure or calculate OTF lumens.

Some descriptions seem a little misleading. I just think it would be great if all the manufacturers would disclose exacly how they get their ratings.

100 LUMENS! (Note: 100 lumens only at midnite during first minute of operation with optional plutonium batteries and flux capacitor-otherwise 20 lumens)

Would be nice to know these things.

Marduke · Jun 3, 2008

Optical looses for a decent LED light are between 20-30%. Incans are 30-40%. General rule of thumb is to just multiple by 0.7, and you're in the ballpark.

Unless otherwise stated, rating are usually emitter. Most well known lights have been reviewed here and we have a good idea of both values.

enLIGHTenment · Jun 3, 2008

Most manufacturers lie about lumen output. Knowing their "measurement" "methodology" wouldn't help unless they actually took measurements in the first place.

Jarl · Jun 3, 2008

Highbeam101 said:
100 LUMENS! (Note: 100 lumens only at midnite during first minute of operation with optional plutonium batteries and flux capacitor-otherwise 20 lumens)

You forgot about the virgin sacrificing also necessary....

Marduke · Jun 3, 2008

Most don't "lie" about lumen output, they quote manufacturer minimum spec emitter lumens. They know the voltage and currently they are suppling to the LED, and they can look on the curve to see what lumen output that corresponds to for the brand and bin they are using, and that's what they advertise.

As an informed flashaholic consumer, it is important to know the difference between OTF and emitter lumens, regulated runtime, runtime to 50% initial brightness (should be nearly the same for regulated lights), and "useful" runtime (which some companies take down to 1-2% of initial output, ie VERY dim moon mode).

There are very few companies who outright lie about their specs.

It is also helpful to know which major companies use which practices.

Keep in mind that usually only very large companies can even afford to measure OTF lumens.

LED-holic · Jun 3, 2008

Thanks for the info Marduke!

Brigadier · Jun 3, 2008

Anyone know what the manufacturing tolerance on LED output usually is? If emitter A is rated at 200 lumens with X number of amps, what is the variation from unit to unit? 1%, 5%, 10%?

KeyGrip · Jun 3, 2008

Good question. Would the manufacturers publish something like that? On the lumens front, I'd prefer a manufacturer to show accurate runtime graphs for the lights. Knowing the exact output is nice, and I support makers that measure out the front, but for my use I'd rather know the runtimes first. Luckily we have some members here who are excellent at making these graphs.

Marduke · Jun 3, 2008

Brigadier said:
Anyone know what the manufacturing tolerance on LED output usually is? If emitter A is rated at 200 lumens with X number of amps, what is the variation from unit to unit? 1%, 5%, 10%?

LED's are binned between a max and min. Tint is done the same way (in 2 dimmensions). A fairly complete rundown is here:

https://www.candlepowerforums.com/threads/156772

A Cree Q5 for example, when driven at 350mA, is guaranteed to emit between 107 and 114 lumens. An R2 bin is between 114 and 122 lumens.

Daekar · Jun 3, 2008

Marduke said:
Most don't "lie" about lumen output, they quote manufacturer minimum spec emitter lumens. They know the voltage and currently they are suppling to the LED, and they can look on the curve to see what lumen output that corresponds to for the brand and bin they are using, and that's what they advertise.

As an informed flashaholic consumer, it is important to know the difference between OTF and emitter lumens, regulated runtime, runtime to 50% initial brightness (should be nearly the same for regulated lights), and "useful" runtime (which some companies take down to 1-2% of initial output, ie VERY dim moon mode).

There are very few companies who outright lie about their specs.

It is also helpful to know which major companies use which practices.

Keep in mind that usually only very large companies can even afford to measure OTF lumens.

I think you're mostly correct, particularly about the knowledge necessary to be an informed consumer, but you (and the manufacturers) did/do leave out decreasing luminous efficiency as junction temperature increases - ie, if the emitter gets warmer, it emits less light per watt. Small lights particularly suffer from this phenomenon AFAIK, and so after a few minutes runtime the circuit might be running the same but light output can drop.

I disagree that only very large companies can afford to measure OTF lumens. Peter Gransee routinely measures the output of not only Arc lights but also of the competition in a NIST traceable integrating sphere, and Arc isn't big. Zebralight also rates OTF lumens for their lights. There are a few more I can't remember. If those two companies can do it, then so can Lumapower, Fenix, EdgeTac, and others on the same scale.

Highbeam101 · Jun 3, 2008

This is great. I'm going to be able to learn more from this one thread and it's links than I have for the last couple of months of trying to figure the rating system out. Thanks guys. Keep it up.

Marduke · Jun 3, 2008

Daekar said:
I think you're mostly correct, particularly about the knowledge necessary to be an informed consumer, but you (and the manufacturers) did/do leave out decreasing luminous efficiency as junction temperature increases - ie, if the emitter gets warmer, it emits less light per watt. Small lights particularly suffer from this phenomenon AFAIK, and so after a few minutes runtime the circuit might be running the same but light output can drop.

I disagree that only very large companies can afford to measure OTF lumens. Peter Gransee routinely measures the output of not only Arc lights but also of the competition in a NIST traceable integrating sphere, and Arc isn't big. Zebralight also rates OTF lumens for their lights. There are a few more I can't remember. If those two companies can do it, then so can Lumapower, Fenix, EdgeTac, and others on the same scale.

The problem is buying the equipment. The initial purchase cost of an IS is huge for most small companies (Arc is part of a much larger company).

However, some smaller companies are starting to outsource IS testing. although it may cost them several hundred dollars per test. I know that EdgeTac had the NiteCore Defender Infinity tested in an IS to 130 OTF lumens (approximately 180 emitter lumens). I was not aware that ZebraLight also rated OTF lumens, but I highly suspect they did the same (outsourced the test).

Maybe over the next year we'll see more IS testing, and clear advertisement when a company decides to do it.

PS

I left thermal drop out because I think that's getting a little too involved for this particular thread IMO. There are just so many variables at play, including the LED brand, how well it is heatsinked, how it is attached to said heatsink, can the light itself dissipate heat, what drive current you're talking about, etc.

It's kinda leaving "flashlight land" and entering "semiconductor world"

Daekar · Jun 3, 2008

Marduke said:
The problem is buying the equipment. The initial purchase cost of an IS is huge for most small companies (Arc is part of a much larger company).

However, some smaller companies are starting to outsource IS testing. although it may cost them several hundred dollars per test. I know that EdgeTac had the NiteCore Defender Infinity tested in an IS to 130 OTF lumens (approximately 180 emitter lumens). I was not aware that ZebraLight also rated OTF lumens, but I highly suspect they did the same (outsourced the test).

Maybe over the next year we'll see more IS testing, and clear advertisement when a company decides to do it.

PS

I left thermal drop out because I think that's getting a little too involved for this particular thread IMO. There are just so many variables at play, including the LED brand, how well it is heatsinked, how it is attached to said heatsink, can the light itself dissipate heat, what drive current you're talking about, etc.

It's kinda leaving "flashlight land" and entering "semiconductor world"

I think you're probably right on all counts. I don't remember if Arc or the parent company owns the IS, but I'd be floored if Zebralight and everybody else actually bought their own when there are labs that will do the testing for you.

I have trouble separating "semiconductor world" and "flashlight land"... all of these topics run together for me.

You're right though, it's probably less significant in a well-designed light than the other factors mentioned.

matt0 · Jun 3, 2008

Just out of curiosity, anyone know approximately how much an IS would cost?

What is Arc's parent company?

Marduke · Jun 3, 2008

matt0 said:
Just out of curiosity, anyone know approximately how much an IS would cost?

What is Arc's parent company?

CIS - Cupp's Industrial Supply
http://www.cpfmarketplace.com/mp/showpost.php?p=2144398&postcount=12

An integrating sphere costs quite a bit, the cheapest, smallest ones being several thousand dollars. One with a suitable size to handle most flashlights with reasonable accuracy would be closer to around $20,000. One on the level of what SureFire has is several times more than that.

NA8 · Jun 3, 2008

I like run time charts. Absolute brightness values can be a little off but you get the big picture of the light output vs time, including initial brightness, fall off with heating in the first 5-20 minutes, quality of regulated run time, and the shape of any moon modes.

Doesn't tell you a lot about the throw vs the size of the spill, tint, quality of the cree rings etc.

enLIGHTenment · Jun 3, 2008

Marduke said:
Most don't "lie" about lumen output, they quote manufacturer minimum spec emitter lumens. They know the voltage and currently they are suppling to the LED, and they can look on the curve to see what lumen output that corresponds to for the brand and bin they are using, and that's what they advertise.

Notably absent from any of that is thermal derating and variance between bins. Not every manufacturer bothers to use the same bin in every production light of a given model. Without including these factors, any emitter lumens estimate is incomplete.

Knowingly publishing incomplete information without disclaiming it as such is lying.

Marduke · Jun 3, 2008

enLIGHTenment said:
Notably absent from any of that is thermal derating and variance between bins. Not every manufacturer bothers to use the same bin in every production light of a given model. Without including these factors, any emitter lumens estimate is incomplete.

Knowingly publishing incomplete information without disclaiming it as such is lying.

See post #10 and #12

Thermal derating is extremely unpredictable and small compared to other factors. You get a larger variance just between supposedly identical circuits.

Most manufacturer specs a minimum bin from their supplier, all lights have at least that bin or higher, and the specs are based off that bin. We have no clue as to if they do indeed attempt to factor in thermal effects or not, but they are not going to put their entire design process on the back of the package for you to read. Asking them to do so is quite silly.

ZebraLight · Jun 3, 2008

We bought an IS last year while working on the H50. It's expensive, but it helped a lot during our R&D process, especially for the H50 lens design. I can understand the the temptation to list a higher emitter lumen figure, but the reflector, lens, and heatsink of a flashlight have so much impact on the out the front figures that we believe only the OTF figure is meaningfull to consumers.

tebore · Jun 3, 2008

Marduke said:
See post #10 and #12

Thermal derating is extremely unpredictable and small compared to other factors. You get a larger variance just between supposedly identical circuits.

Most manufacturer specs a minimum bin from their supplier, all lights have at least that bin or higher, and the specs are based off that bin. We have no clue as to if they do indeed attempt to factor in thermal effects or not, but they are not going to put their entire design process on the back of the package for you to read. Asking them to do so is quite silly.

Actually the old Lux's had thermal derating in their spec sheets. I should add that in most electronics that specs are rated at 25 degrees Celsius. So it's possible to figure the drop in output as temps go up.

If you notice in all lights where you test for lux you'll see the graph starts off real high then drop then stay steady(regulation) or drop as the battery dies. This initial drop happens in about 5mins.

Some lights have the ability or compensated for this drop caused by thermal. I believe the original HDS mentioned something about it. It's not silly. If you're an engineer you should be able to do it. In the real world there are very few unknowns in products.

At the emitter, out the front. What's the deal?

Newly Enlightened

Flashaholic

Enlightened

Flashlight Enthusiast

Flashaholic

Flashlight Enthusiast

Enlightened

Flashlight Enthusiast

Flashaholic

Enlightened

Newly Enlightened

Flashaholic

Enlightened

Enlightened

Flashaholic

Flashlight Enthusiast

Enlightened

Flashaholic

Enlightened

Flashlight Enthusiast

Similar threads