The Nightsword project

Candle Power Forums

Help Support Candle Power:

In a nutshell, the long FL gives a powerful beam because it places the lamp further from the reflector which produces more throw.

With "throw" you mean sort of practical usability here, not the max lux in the center of the spot, right?
 
Actually, a short FL could potentially produce a higher "peak luminance", but only within the very center of the beam. A long FL produces higher "net luminance", even with slightly less "peak luminance", which makes for a more visible beam overall because nearly all light within its beam is at its peak luminance, rather than just the small amount of light at the very center. I consider "net luminance" to be the true measure of a beam's power.

Here is an outline of why this occurs.

Edit - Note that these comparisons are without a retro-reflector. A long FL with retro-reflector will produce both higher "peak luminance" and "net luminance" than a short FL.
 
Last edited:
Actually, a short FL could potentially produce a higher "peak luminance", but only within the very center of the beam. A long FL produces higher "net luminance", even with slightly less "peak luminance", which makes for a more visible beam overall because nearly all light within its beam is at its peak luminance, rather than just the small amount of light at the very center. I consider "net luminance" to be the true measure of a beam's power.

Here is an outline of why this occurs.

(I guess you mean "peak illuminance".)

How strange, although I do know the linked pictures well, I never noticed that peak illuminance is not the same.

Is this difference mainly a result of your calculation or can you give an explanation?
Is it some practical side effect which I'm currently not thinking of, or does it even conflict with the idea that peak illuminance
is only determined by luminance which remains constant (conservation of etendue), apparent reflector disk area, and losses (absorption, scatter)?
[edit: given that the reflector is good enough to show peak luminance over its entire apparent area, which can be tough for small luminance hot spots]
 
Last edited:
(I guess you mean "peak illuminance".)

Yes, however it could be described as illuminance as it relates to the intensity at a given distance (Lux), or it could also be described as luminous intensity as it relates to the intensity within an angle of the beam (candlepower or candela).

But to help clarify what your really asking, imagine two very different unrelated beams... one that projects 100 Lux within an area of one square meter at 1 km, and another that projects 75 Lux within an area of 4 square meters at 1 km. The first one is brighter within the small area it's projecting within, but the second one is more visible overall because four 75 MCP lights is more visible overall than one 100 MCP light.

What's key here is that even though the 75 Lux by 4 square meter beam is only 3/4 the illuminance of the 100 Lux by 1 square meter beam, the 75 Lux by 4 square meter beam illuminates double the atmosphere thickness of the 100 Lux by 1 square meter beam. Viewing the beams from the side, the 75 Lux by 4 square meter beam has 3 times the atmospheric illuminance of the 100 Lux by 1 square meter beam.

That's what I mean when I always say candlepower isn't everything. "Net Illuminance" is really what determines how bright the beam is.

By the way, these MCP values are off a bit because 1 Candlepower actually equals 0.981 Candela. Incidentally, I just realized all of my MCP values given thus far have been off by almost 2% because I hadn't taken this factor into consideration.
 
I just felt compelled to point out that to me, this thread is a diary of PROGRESS, not documentation of a lack of progress.

You have repeatedly worked to perfect your light, and to over come obstacles that arose as you worked through that process.

If I were an employer looking for an engineer to develop a product, this thread should be on the engineer's resume.

:D
 
That's what I mean when I always say candlepower isn't everything. "Net Illuminance" is really what determines how bright the beam is.

That was not my question. I wanted to know why the very "peak illuminance" differs in your opinion.

I've looked again and now I noticed that not only the focal length is different but the lamps are, too!
So I wonder if the linked pictures/calculations help here. Back to my original question:

Actually, a short FL could potentially produce a higher "peak luminance", but only within the very center of the beam.

Why is that?

Perhaps the two calculations here (#221) better match this question?
(because they show: same light source, aperture and vertex diameter. but different focal length -> why different peak illuminance?)
 
Thank you TEEJ, I definitely tried but was unable to find anything in those last criticisms that could be used to improve this, or even anything that was correct.

Sven, your basic question is what makes the different focal lengths create different beam profiles.

It's explained here, but here is a more thorough explanation...

Collimation (the degree to which the photons project in the same direction) determines throw. There are two fundamental geometric components that determine collimation: the size of the source and the distance of the source from the reflector surface. To clean this up, I'll refer to that distance as STRD (Source to Reflection Distance). Smaller source size and greater STRD produce greater collimation. Refer to this most basic illustration...

strd.png


The beam is a distant reflected "Image" of the "Source" (RED), enlarged at distance based on STRD. The greater the source size and the smaller the STRD, the less collimation occurs and the greater the reflected image size.

The larger image (BLUE outlines) is the result of the shorter STRD.

The smaller image (YELLOW outlines) is the result of the longer STRD.

It's a direct relation...

Double the source size equals double the image size, half the throw, and 1/4 the image intensity.

Double the STRD equals half the image size, double the throw, and quadruple the image intensity.

Likewise, double the source intensity equals double the image intensity.

The final beam is a combination of all images generated from all reflections over the reflector surface.

As this illustration is a short focal length reflector, there is much STRD variance, generating a beam of much image size variance. Hence a small central spot and a large dim corona. The bright central spot is the result of the very large STRD from the source to the very front of the aperture. Although that large distance has great collimation, only a relatively small percentage of the source light actually hits that region of the reflector. The majority of source light has a very short STRD which has very little collimation, making for a very large dim corona.

On the other hand, a long focal length reflector has much more uniform STRD over it's reflection surface, generating a beam of much more uniform size. There is less centralized intensity but less light wasted as large dim corona
.

That's it in a nutshell.

On a side note, there are several other geometric factors occurring which I calculate...

-The intensity of the source is not uniform across all emittance angles. An angular intensity profile is measured and mapped for each source.

-Since the source is not perfectly spherical, the size of the facing area of the source changes at every reflection angle. For HID lights, the beam calculator uses a formula for calculating the angular facing area of an ellipsoid defined by the source width and height.

-The "Image" distance is resolved as from the aperture rather than the source.

-The aperture diameter at each reflection point has a geometric affect on the image.

-Focus offset for flood.

-Vertex hole diameter

Here is a diagram of all of the full geometry calculated for just one reflection point with offset focus:
strd-full.png

As you can see, it takes several geometric functions and laws to evaluate for just one reflection point on the surface of the reflector. This has to be done across the entire surface of the reflector. And this diagram is without a retro-reflector. Of course there are also non-geometric factors calculated including source lumen output, reflector reflectance, and lens transmittance.

EDIT - I forgot to mention something I discovered a few days back that others might find fun.. due to the beam uniformity as a result of uniform STRD, and the direction of focus offset, a metal cutout "light shield" could be used for the Nightsword to project an image like the batman logo in the clouds by slightly defocusing the beam. I thought this would be fun.
 
Last edited:
Sven, your basic question is what makes the different focal lengths create different beam profiles.

Is it? Please don't think that I might underestimate the importance of the beam profile.
But here I'm interested in a special, yet fundamental point. I can't imagine yet, how peak illuminance might depend on focal length.

Thanks for the detailed explanation, especially about some of the background how your calculator works.

I'm afraid I don't get the connection of "STRD" to peak illuminance, yet.
But the following made me listen up:

The intensity of the source is not uniform across all emittance angles. An angular intensity profile is measured and mapped for each source.

Do you have a uniform source available for your calculator? (the listed sources don't seem to be)
And could you please run once with such a uniform source and different focal lengths?

Sure, this might perhaps not be realistic for any HID. But LEDs w/o dome might get pretty near to ideal here
(although only in one half-space, certainly)
 
Last edited:
I just bolded what you need to understand in order to see the answer to what you're asking.

EDIT - LEDS are not any more uniform in angular intensity than HID. For example, see the first diagram on page 7 of the Cree XM-L2 data sheet. Additionally, the LED emittance surface is a half ellipsoidal shape, but the calculator calculates for full ellipsoidal only. As mentioned before, it would take a lot of work to arrive at the formulas to calculate this and I'm not interested in LED.
 
Last edited:
I just bolded what you need to understand in order to see the answer to what you're asking.

Sorry, I don't see how it explains peak illuminance. The STRD effect still could perfectly balance for the target illuminance at a very point.
that is, these effects from your quote might just balance:
"large STRD distance: great collimation, small percentage of source light hits reflector"
"small STRD distance: little collimation, large percentage of source light hits reflector"

I am not talking about the whole beam profile here, or "net illuminance", or how bright the beam as whole appears.
Please note I am completely with you about the effect of focal length on the beam profile. But not peak illuminance.
I'm curious if you believe peak illuminance dependency from focal length can only be determined by such an overall calculation
or if it even can be really understood.

Why am I asking this and probably bothering you (I hope not)?
As I understand it, peak luminance must not depend on focal length for a source with constant luminance.
And if you don't have a beam calculator available, this becomes a pretty fundamental question for calculating
max lux of a thrower, and thus can be helpful for rough estimations.

EDIT - LEDS are not any more uniform in angular intensity than HID. For example, see the first diagram on page 7 of the Cree XM-L2 data sheet.

Oh, I meant uniform, constant luminance of the source. The effect that luminance doesn't depend on the angle
you look at an LED without dome. But perhaps that was a misleading comparison. Let's forget about the LED.

Do you have a source with constant luminance available for your beam calculator?
 
Even de-domed, LED emits only in one direction, not the back side, and the luminance shape is then a flat square, so would still take much time to arrive at formulas for and wouldn't help you with what I think you're tying to find out.

I think what you're trying to discover is if there is a fundamental geometric relation of focal length to peak illuminance alone, and a spherical source of uniform angular intensity would help narrow the factors to maybe recognize a peak illuminance pattern associated with focal length alone. The fundamental geometric relation is the "source to reflection distance", but it would be interesting if the relation extended beyond that in a manner I can't yet conceive of. I will create a uniform pattern to play with but my guess is it's unrelated other than which happens to maximize the source to reflection distances.
 
(Yes, please just forget the LED. I'm tempted to clear up that I actually meant something quite different, but it's not important.)

An illustration might help to express my point. See the last picture in this post from Ra below.

If focal length or STRD had an influence on the max lux, then I would really expect to see
some effect of this in the reflector if you look back from the target to it.
Although the picture certainly shows quite some variations: darker areas where the
unperfect reflector just doesn't catch the light from the almost point like sources,
I believe that surface brightness at the outer edge is the same like near the vertex hole.

And "conservation of etendue" also can be interpreted that focal length doesn't count.

Now, if focal length becomes relevant for max lux nevertheless, in some practical situations,
and in a way that can be understood (that is, without a calculation program only), then I'm curious like hell.
I hope you see why I jumped in at your remark.

I've modded two Thor spotlights: One with 35/50watt HID, and one with Mercury-short arc (super high surface brightness!)

I placed both spotlights in the field and photographed the difference in apparent surface brightness: (forget the smaller lamp in the middle:that is Maxabeam)...
surfacebright7hf.jpg
 
Last edited:
I've created the uniform luminance profile and tested with a sphere shaped source by entering matching width and height for the source size, and entering a zero reflector vertex hole diameter. As anticipated there's not a fundamental relation of peak illuminance to focal length alone. Either long or short FL can produce more peak illuminance. It's simply which FL gets the source further from the reflector surface. The long FL gets the source further by locating it further forward of the reflector, and the short FL gets the source further by locating it further inside the reflector in which the forward-most reflector surface at the aperture is furthest from the source. And medium FL has least peak illuminance because it's smack in the middle of the reflector with no regions of long "source to reflector distances".

Note that the regions of greater surface intensities of Ra's reflectors will change depending on how far away the pic is taken. If you were to view the short FL reflector "straight on" from a great distance, the outer edge would be much much brighter than the rest of the reflector, but if you viewed it up close, the inner edge would be much brighter. This is because the inner edge has a lot more source light hitting it, but it's very poorly collimated, and the outer edge has little light hitting it but it's highly collimated. A short FL reflector can produce an extremely intense beam up close because the vast majority of the light is reflecting from closer to the center of the reflector, but it quickly fans out with just the outer edge of the reflector collimating well.

Here is something interesting I'm noticing with these calculations.. with a uniform luminance source, both peak illuminance and net illuminance differ by the same degree when changing the focal length. That is to say, a peak illuminance always corresponds with net illuminance. Maybe this is what you were getting at. Of course this is only for when using a uniform luminance source, which there are none, and with no other variables involved. In the real world with non-uniform luminance sources and the additional variables of the reflector vertex hole and non-spherical source area, peak illuminance doesn't correspond with with net illuminance and it is possible to have one light with more peak illuminance and another with more net illuminance. Just as a laser has greater peak illuminance than a very large search light, the laser beam is less visible from the side at great distance because it has less net illuminance.
 
Sven, here's a comparison of three configurations all with the same lamp and same aperture, lens, same everything, even the same vertex hole diameter, no retro-reflectors, just with different focal lengths... a very short .5" focal length, a long 3" focal length, and a medium 1.5" focal length.

The very short focal length has the highest peak illuminance, but the long focal length actually has slightly more net illuminance even while having 58% less peak illuminance. The medium focal length has much less of both peak illuminance and net illuminance, as a result of the source being right in the middle of the reflector with no long STRD. It's worth noting that the long focal length occupies 1/6th the space of the very short focal length, packing a way bigger punch for it's size.

All three rendered with their peak illuminance relative to 107 Lux...

net-illuminance-comp-2.png


net-illuminance-comp-1.png


net-illuminance-comp-3.png


As mentioned previously, there is one additional factor which this doesn't take into account for. At lower and flatter incident reflection angles on the reflector surface, comes a greater degree of light loss and accuracy loss due to the nature of reflective coatings, which somewhat hinders short focal length configurations. This doesn't affect long focal length because the reflection angles are from straight on perpendicular to no less than 45 degrees.

EDIT - Got tracking on the new sample touch panels. Last night I couldn't help but build a new updated universal-orientation weather proof air intake assembly to test and compare it against the limited-orientation assembly. I played a bit with it while being just half complete (and half effective), and it's looking to be very promising. I'll finish it up tonight.

Also note that candlepower is no longer exactly Lux @ 1 km. I've updated the program to properly evaluate Candlepower as 0.981 Candela.
 
Last edited:
I like your first example with uniform light source.
And I guess we should stay with such a constant luminance source for testing the influence of focal length/STRD on peak illuminance.
The mercury lamp would complicate it.

And the Maxabeam has a good reflector example: It has a short focal length, and the distance of the light source to the mirror (STRD) varies heavily.
- outer diameter: 118mm diameter
- vertex hole diameter: 10mm
- vertex focal length: ~9mm.

Now I have a thought experiment in mind, which emphasizes on possible STRD effects and can be easily fed to your calculator.
And as it is a real reflector, it could be compared to experiments.

Lets take two parts of the reflector, a small ring near the center hole, and a small ring at the outer edge.
Both rings should have the same apparent area (to ease further calculations a bit). But STRD would be extremely different.

These would do it:
1st ring: inner diameter 10mm (vertex hole), outer diameter: 14mm, -> 75.4mm²
2nd ring: outer diameter: 118mm (edge), inner diameter 117.5925mm, -> 75.4mm²

Now let's look back from the spot to these reflectors:
- if the light source appears with the same brightness in both, then max-lux should be the same (as the area is the same)
- if STRD or focal length have an effect, the light source would appear with different brightness, and thus max-lux would be different

So, if you feed this to your beam calculator, different max-lux values would mean that the rings would have to appear with different brightness.

Do you agree?
 
Of course, but dependent upon distance measured because up close the inner is more intense due to much greater light gather, but at distance the outer is more intense due to much greater collimation. This is explained in the 2nd paragraph in post#1053.

It's evident in the geometry without a thought experiment, but for the fun of it I've performed such tests with real reflectors by taping sections of them off with electrical tape, and I've also toyed a lot with this in the past using the calculator.

I was unable to use the values you suggested because that 2nd ring is not thick enough to full-fill an iteration of the calculator (each iteration is one angle of the source), but using equivalent ring thickness the inner is 5 times brighter right up close as a result of having 5 times more light gather than the outer ring, but the outer is 14 times brighter at 1 km even with 1/5th the light gather of the inner ring.

EDIT - The inner surface contributes to the candlepower at distance by only a fraction of 1/14th compared to the outer surface. The entire inner surface is essentially useless for throw.

That's why I would use a forward facing retro-reflector if I were to build a short FL light, it would redirect the rear-ward light onto the forward surface of the reflector to benefit by it's 14x collimation. The beam calculator can be used to determine precisely at which source angle the retro-reflector would become most effective.

If you really want to make the most direct comparison, we would use equivalent ring thickness when viewing the reflector from straight on, in which case the outer ring would fare even better, but this is close enough to get the idea.
 
Last edited:
Of course, but dependent upon distance measured because up close the inner is more intense due to much greater light gather, but at distance the outer is more intense due to much greater collimation.

Just use a long distance, otherwise we had focus problems anyway.

If you really want to make the most direct comparison, we would use equivalent ring thickness when viewing the reflector from straight on, in which case the outer ring would fare even better, but this is close enough to get the idea.

Of course the outer ring would fare much better, if you take equivalent ring thickness: The area of the outer ring is much higher!
The point I'm after is rather the "apparent brightness" of this ring (the luminance) if you look back to the reflector.

(That's why I chose equivalent areas: Same resulting target-lux means same apparent brightness in reflector.
But if you do need to take a greater area for the outer ring then you have to divide resulting lux through the factor this area is higher.)

I'd like to show this: if focal distance or STRD determine target lux, then you have to see this if you look back to the reflector.
Inner zones would have to appear with different "apparent brightness per area" (luminance) than outer zones.
And if there's a high range of the distance of light source to reflector surface (like with the deep maxabeam reflector, from ~9mm to about 80mm)
then there should be quite some range of apparent brightness.

Do you agree here?
 
To compare "apparent" brightness of the outer ring to the inner ring when facing the reflector directly on the front, compare with equal "apparent" ring thickness also when facing the reflector directly on the front.

EDIT - revising the results of water testing...
 
Last edited:
I've completed water resistance testing/comparing the new universal-orientation air intake to the limited-orientation. There are four tests:


1. High Flow Shower Head Spraying Downward on the Intake

Limited Orientation Result: intake can be angled in any orientation except "beyond momentarily" in upside down angles beyond straight up.

Full Orientation Result: intake can be angled in any orientation without limitation.


2. High Flow Shower Head Spraying All Directions on the Intake

Limited Orientation Result: intake can be angled in any orientation except "beyond momentarily" in upside down angles beyond straight up.

Full Orientation Result: intake can be angled in any orientation without limitation.


3. Full Water Force of VERY High Flow/Volume Tub Faucet Pouring Downward on the Intake

Limited Orientation Result: intake can be angled in any orientation except in upside down angles beyond straight up (not even momentarily).

Full Orientation Result: intake can be angled in any orientation except pointing beyond 45 degrees downward, but only lets in a few drops which is harmless.


4. Full Water Force of VERY High Flow/Volume Tub Faucet Pouring all Directions at the Intake


Limited Orientation Result: intake can be angled in any orientation except in upside down angles beyond straight up (not even momentarily).

Full Orientation Result: intake can be angled in any orientation except pointing beyond 45 degrees downward, but only lets in a few drops which is harmless.


Now I'm preferring full orientation. Does your preference change BVH?
 
Last edited:
I'm a little confused on the #2 "Fully weather resistant and direct water pour resistance under 180-degree (up to down) orientation" option back ending with post 975 and now the wording "limited Orientation" and "Full Orientation". Please clarify for me.

But for me, it comes down to:

What are the cosmetic/appearance differences between Limited and Full? - If no difference, then "full"
What are the cost and complexity (to build and to service) differences between the two? - If minor increase, then "full". If significant, then possibly "Limited".
 
Back
Top