How to "eyeball" lumens?

[musicalfruit]

I see some messages with "eyeball" estimates of lumens and was wondering if there was any tips to doing it.

Or is building some type of rig like Quickbeam's the "proper" way to do it short of access to an integrating sphere?

 

[Flying Turtle]

For me the best way to "eyeball" the differences between two or more lights is to do the "ceiling bounce" test. I look straight ahead and alternate reflecting the lights off the ceiling. Usually I'll hold the light over my head to avoid seeing it out of the corner of my eye. It's pretty evident which light is putting out more lumens this way.

Geoff

 

[FlashCrazy]

That's how I do it. A small room with light colored or white walls works best. I use my bathroom....works great, since any substandard lights can be easily flushed down the toilet.


Kidding...I'm kidding!...

 

[KingGlamis]

I use the bounce test and outside tests as well. But has anyone noticed that the longer you own a light the dimmer it seems, even with fresh batteries? I guess I just get used to a light and am not so impressed after I have owned it for a while... then I want something BRIGHTER! :laughing: I guess it's kind of like buying a fast car. It seems REALLY fast the first week or so that you own it. Then it gradually feels slower and slower as you become used to the power.

 

[LED61]

I too use ceiling bounce tests but comparing known lumens of other flashlights is a must.

 

[FlashCrazy]

Yeah, getting used to it plays a big part on why they don't seem as bright after time. Another factor may be LED degredation, which has been discussed on the forum. It seems to be more of a factor on hard driven LEDs. I haven't really noticed this, but I guess it could play a part.

 

[musicalfruit]

So the ceiling bounce is a good way to compare output between two lights.

But how would you quantify the difference? If one light looks slightly brighter than your reference 100 lumen light, is there a way to measure that it's 20 lumens brighter or 50 lumens brighter?

 

[KeyGrip]

But how would you quantify the difference? If one light looks slightly brighter than your reference 100 lumen light, is there a way to measure that it's 20 lumens brighter or 50 lumens brighter?

That's tricky because of the way our eyes percieve light, especially differences in light levels. Giving relative figures (light x is brighter than light y) seems better than trying to give quantitative values ( light x looks at least 30 lumens brighter than light y)

 

[cy]

totally agree ceiling bounce test is good to compare side by side output for different lights.

but to eyeball lumens... it takes good old fashion using the light.

best to do this with lights with known outputs. for instance unmodded ARC LSH-P puts out aprox. 25 lumens. PD w/luxeon puts out aprox. 45 lumens and so on...

after a while you be able to tell if light is performing similar to others.

 

[cy]

totally agree ceiling bounce test is good to compare side by side output for different lights.

but to eyeball lumens... it takes good old fashion using the light.

best to do this with lights with known outputs. for instance unmodded ARC LSH-P puts out aprox. 25 lumens. PD w/luxeon puts out aprox. 45 lumens and so on...

after a while you be able to tell if light is performing similar to others. realizing the best you can do without an integrating sphere is get it in the ballpark.

 

[Windscale]

I do bounce tests as well. But I also measure output using my Sekonic light meter (measuring incident light). But it is necessary to take a few readings and at different distances. I do 3 feet, 10 feet and 20 feet. Take a few readings with each distance to make sure the light is pointed directly to the incident light receptor. Moving the light very slightly will result in a different reading. The highest reading in each distance counts. The results are in EVs. I am not sure how to convert these to lumens. But these can be compared to manufacturers' advertised lumens for references. I just keep my data in EVs.

 

[robm]

Direct beam light meter readings are equivalent/can be converted to the lux or throw values of the light. You can't directly measure lumens/output with this type of reading.

Although for lights with similar beams the 2 are often related.

 

[Codeman]

I've found quickbeam's setup to be accurate within 5-7%, certainly good enough for me. To get that level of accuracy you have to be very consistent with the setup and use several lights of known lumen output to arrive at a meaningful conversion factor for a particular setup. And it's only useful for white lights. Colored LED and such aren't measured accurately by common light meters, as the CPF benchmarking lights have shown.

I'm not a big fan of ceiling bounces as a way to estimate lumen output. While it appears to be a good method, and it may very well be for some folks, I've found that it is prone to unintentional subjective bias. There's been several times where I thought 2 lights related to each other in a certain way, only to have my light box/meter prove it to be otherwise.

Not many folks using quickbeam's approach...which is too bad, IMO.

 

[robm]

...Not many folks using quickbeam's approach...which is too bad, IMO.

I do, and have also found it much more consistent than 'ceiling bounce'.

Also, as I am lazy, I can compare different light configurations (lamps, batteries, switches etc), without having to get up to go to my 'white-ceiling-bounce-test-room', - switch on light, point through hole, read light meter, make a note of reading (that last bit is quite useful to remember)

However - ceiling bounce is still required for larger lights that don't fit the hole in the light box.