The LightBox Experiments... Comments please!

[brightnorm]

[ QUOTE ]
PaulW said:
...As you probably already know, the more floodlike the beam, the more repeatable the measurement and the less sensitive it is to flashlight orientation. For flashlights with narrow beamwidth (long throw), I suspect orientation is the biggest problem. That is, the measurement varies wildly as the light is moved. I would guess that putting additional baffles in the box would reduce this effect. Ideally, it would be nice to be able to move the flashlight to any angle and still get the same measurement...

[/ QUOTE ]

Paul,

You can minimize that effect through a curved (ideally sperical) reflector and a positioning jig. I mentioned in my post that I was able to get virtually identical results between the Stinger and the Stinger HP, two totally disimilar beams of equal total output.

One idea for a poor mans integrating sphere:

Get a cheap plastic world globe. Poke holes for lens and meter, paint the outside white, check to make sure interior is white. If interior isn't white then spray paint and cut hole on opposite side to finish spraying.

Brightnorm

 

[brightnorm]

[ QUOTE ]
BlindedByTheLite said:
hmm it seems from everything i've heard about the peli m6 that reading's kinda low.. i always thought it was noticeably brighter than the scorpion and the a2...

[/ QUOTE ]


The PM6 is "brighter" than most other similar lights because of its narrow beam and efficient reflector. Quickbeam was measuring total output as opposed to focused brightness. When SureFire issues lumen ratings they are measured in an integrating (or integrated) sphere which measures ALL the light being emitted. This shouldn't be confused with "brightness" (candlepower) which is a measure of the brightest spot in a beam.

A 50 lumen light can have greater brightness (candlepower) than a 500 lumen light if you narrow the small light's beam to, say 1/2" diameter at 5 feet, and widen the big lights beam to 50' at 5 feet. That way, any given point in the big beam will be dimmer than the little beam, which only proves that you have to be alert when reading ads about brightness.

Brightnorm

 

[BlindedByTheLite]

thank you for clearing that up for me brightnorm.. that sounds very familiar now.. almost.. deja vu like..

 

[RevDavid]

If you put a diffusion baffle between the part of the box with the flashlight, and part with the meter, it may cut down on the directionality problem.

You might also be able to feed the light into the box through either a white snoot or a fiber optic, either of which could be at a fixed angle to the inegrating box, and that might be even more effective at eliminating the problem.

I do think that a small yellowish color shift in the box will affect your readings on different colors of lights. It will have more effect on blue than green or red, and more on white LED than incandescant. The strength of the effect will vary by spectrum. The biggest problem you will face, in my opinion, is that the "whiteness" of your box will vary over time, as the paper yellows. You will get slightly more consistent and representative results if you can get a whiter surface that resists yellowing with age. I think that might be why Omega used white styrofoam in their color heads. I have seen them at different ages, and I don't remember ever seeing one that had gone to an ivory tint.

If you were measuring light in something like f-stops, this would be a moot point. But your readings are a lot more sensitive, defining very small differences in output. I think even the difference between two new batteries might show up. You could check that easily enough with a single cell light. Anything that keeps the readings more consistent would actually show up in your tests.

Keep it up. This is an outstanding project!

David <><

 

[shiftd]

i just saw this thread and IMHO, that is the best idea that us, moneyless human can come up with regarding the lumens reading /ubbthreads/images/graemlins/tongue.gif

great job, quickie /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/thumbsup.gif

 

[MR Bulk]

Doug,

This is great -- now take off all optics/reflectors from your lights and re-test with bare bulbs/LEDs -- and see if the relative numbers are still proportional. If so, you got a good thing going here...!

 

[PaulW]

Taking the reflectors off some lights will change their optical properties substantially. For example, I'm thinking of a Mag 3D host I have for an incandescent. When I take the reflector out, I have only a bare bulb in a dark red cavity -- so dark that I think it might absorb well more than 50 per cent of the light.

Paul

 

[Quickbeam]

WOW! Thanks for all the input! First of all:

<font color="blue">"This is great -- now take off all optics/reflectors from your lights and re-test with bare bulbs/LEDs -- and see if the relative numbers are still proportional. If so, you got a good thing going here...! "</font>

The only problem I see with this is that it doesn't take into effect any light loss due to the lens/reflector absorbing some light - hence the testing of the complete light. Some crappy lenses may absorb a lot more light than pyrex, for example.

<font color="blue">"i just saw this thread and IMHO, that is the best idea that us, moneyless human can come up with regarding the lumens reading"</font>

"Moneyless" being the key word! Plus anyone can reproduce it very easily with their own meter.

<font color="blue">" The biggest problem you will face, in my opinion, is that the "whiteness" of your box will vary over time, as the paper yellows. "</font>

Luckily I can throw out the box and get a new one anytime! 1/2 gallon of milk later and I'm ready to go. Just cut the holes! That's the great thing about this particular setup.

Brightnorm/Blinded:

You describe the exact problem I'm trying to solve by using only beam Lux measurements to describe the light's ouput. Optimally we'll have 2 readings for each light - Lux at beam center (throw) and Lux from the lightbox (Overall output). This will help describe the output of a light a bit better. High throw # and low Overall # = tight beam from dim element. Low throw # and high Overall # = bright wide angle beam.

Perhaps a comparison between the two numbers in order to describe the beam e.g. 2000/50 = 40, 200/5 = 40 means both have similar beam angle vs. output despite different intensities... Hmmmmmm... I'll have to put some thought into this...

Brightnorm: /ubbthreads/images/graemlins/ohgeez.gif I shoulda thought of the cheap globe idea! Too late now, I'm going to keep going with what I've got for now. And thanks for checking my numbers against your setup - It's good to hear that the numbers correlate - That means that the data from this project is somewhat valid!

And Last... FNG: Sorry - no 3 cell surefire lamps /ubbthreads/images/graemlins/frown.gif

More numbers will be coming soon!

 

[MR Bulk]

PaulW,

A-hah! My point precisely. As long as the bulbs/LEDs can be exposed 360° and with their relative positioning a minimum of at, or above, the plane of the base/socket joint (meaning on the Mags you'd have to remove the head as well as the reflector, with the base of the bulb above the rim of the battery tube), I think we'd have a truly level playing field with which to make comparisons.

The Space Needle II, for example, lumen'd 216 in a major flashlight mfr's i/sphere, and yet with the head and reflector put back on and the highly concentrated, tightly focused hotspot of a beam pointed straight at a lux meter, it will read in the multiples of thousands. I know lux and lumen are far from interchangeable, but I'm just trying to provide some contrast here. In fact -- I should try and lux-meter the light without head or reflector as well...

 

[yclo]

Something to do with my new Lux meter.... /ubbthreads/images/graemlins/smile.gif

 

[Quickbeam]

Here's a couple of quick comparion numbers for everyone:

Minimag: 580 (not much more than the A2 LEDs!)
3-D Mag: 1520 (less than the Streamlight 1L incandescent!)

(chart and list above updated.)

Both with stock bulbs and lenses.

All of these tests are being done with very new batteries.

<font color="blue">"As long as the bulbs/LEDs can be exposed 360° and with their relative positioning a minimum of at, or above, the plane of the base/socket joint (meaning on the Mags you'd have to remove the head as well as the reflector, with the base of the bulb above the rim of the battery tube), I think we'd have a truly level playing field with which to make comparisons."</font>

Perhaps I'm missing a point here, but I just don't see how removing the reflector and optics would give any sort of meaningful reading. With my setup you are measuring darn near all of the light as projected out of the end of the light. This is what the end user sees, not all of the light projected by the bare bulb. Since some light is lost before it gets out of the end of the flashlight, reading just the bare bulb could give readings that are higher than the actual output the user sees.

Not that my setup has a high degree of precision or anything, but I'd prefer not to add anything that could skew the numbers any further. Fact of the matter is that the playing field is not level due to the differences in materials used to project the light and the light losses as a result of those differing materials.

 

[AilSnail]

What MrBulk proposes would give some indication of the relative efficiency of the reflector and lens. Also, if power was known, it would give an indication of the bulb's efficiency.

Looking forward to the numbers comparing relative lumens of the diffusing films, for instance, not to mention stochastic reflectors. Another interesting thing would be to see if focus on ie mags and pellicanes affects lumens.

 

[Quickbeam]

Re: The LightBox Experiments... COMPLETE!

...AND HERE IT IS:

http://flashlightreviews.home.att.net/reviews/lightbox_output.htm

Added about 30 more lights!!

 

[TOB9595]

Thanks very much, QB /ubbthreads/images/graemlins/smile.gif
The gif's are mighty handy
Tom

 

[4x4Dragon]

i havent read through this whole post but i've noticed that the first 2 numbers of the reading for each light seem to come real close to the advertised Lumens output of the lights I have/had! Good job! it's at least a good basis to compare lights!

The poorman's Lumensphere!

 

[JohnK]

Hmmm. I think 4X4 is right, a fair approximation of lumens it is.

This is going to be a VERY useful tool.

 

[Doug S]

[ QUOTE ]
Quickbeam said:

Comments? Shall I continue?

[/ QUOTE ]

Hell yes! QB Doug, this rates a /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/thumbsup.gif in my book. I have been waiting for someone to tackle this problem and have thought that either you or Roy where the best ones to do it.
It is clear to me that you have studied integrating sphere design and have tried to address the basic design essentials. Here are my comments. I think that you have already thought about everything I am going to say but you asked for comments so here they are. For relative comparisons between lights of differing spectra such as White LED vs incandescent or colored LED the "whiteness" of the surface does matter. It may be that your brand of milk carton is very good it this regard but it is hard to know for sure. While I don't know the how consistant the reflectivity of styrofoam across the visible spectrum is, I like RevDaves suggestion of styrofoam on an intuitive basis as it seems very pure white to me.
You mentioned a +/-5% relative reading depending on the relative focus of the light. Are we to assume that this was a test performed with a focusable light such as a Mag? If so, this is a good test of how well your box is integrating. As we know, a sphere is optimal but if you are getting +/-5% in a rectangular box, this is pretty darn good. I believe [going slightly out on a limb here just because I am too lazy to go back and look at the equations] that with any reasonable enclosure geometry, the quality of your integration improves with the reflectivity of the interior surface. If styrofoam or some other surface is significantly brighter [more reflective] than your brand of milk carton it would be worth a try. Another thing to be aware of is that the error introduced by differences [in reflectivity] across the visual spectrum can be larger than what might seem intuitive. For example, if the white surface is 10% more reflective of red than blue, the relative measurement obtained for a red light can be >> greater than 10% higher than a blue light where each are actually equal. The size of this increased error increases as the reflectivity approaches 100% for one of the colors.
When replicating your test setup, be aware that small changes in geometry [for example, the hole sizes] can have large changes in results. The same is true for small changes in reflectivity. If the dog eats your milk carton and you have to build a new one, it would be a good idea to calibrate your next box using some "archived" light that you tested in the previous box.

You da man /ubbthreads/images/graemlins/thumbsup.gif

 

[JohnK]

Doug, I agree with you. My earlier statements were just that. A tad grey/brown, whatever color of the carton lining could alter readings a lot. Even ageing will change color.

But never the less, this is a really great tool, and I much prefer it to Lux readings.

 

[radellaf]

Bravo! Glad you went through and did it. Maybe styro would be better but at least you have a good start. Any Brinkmann LX around to test?

I find the 1000/1500 levels for the 2D & 3D lights revealing. So much for Sears's claim that the 3D tool light is "200% brighter" than the 2D light. I thought the 3.6V lamp is more efficient than the 2.4V one, but not that much more. I'd love to see if the Mag-Num Star Xenons Home Depot's selling put out significantly more light than the Kryptons. What's really odd is they draw more current, initially, yet Mag rates the run-time as 15-16 hrs instead of 10-12 (Kr). Maybe they still work at lower voltages than the Kr. Anyway, I digress..

 

[Quickbeam]

OK, OK... I know it may not be the most accurate device in the world... There may be an issue with wavelengths, etc... But on a budget of $5, whatdya expect?! /ubbthreads/images/graemlins/icon15.gif

Styrofoam, eh? I wonder if one of those cheapie styrofoam coolers would work better. I'd hate to do all of the readings over again so soon, but I'll keep an eye out next time I go shopping. It would be easy enough to make work... Hmmmmmm......

Yes. I'll look into it. /ubbthreads/images/graemlins/smile.gif

I'm still pondering the cheap globe painted white inside too...