Throw - Incan vs LED ?

Sometime back I read that with a similar sized reflector and the same lumen output a Incan light will out throw an LED. If this is true - Why? I believe the post also said that in rain or fog an Incan is also superior to an LED light?

I'm guessing it would have to do the the point source of an incan. I don't actually know that much about incan's but I'm guessing maybe the light comes from a smaller point (source) than a led.

If that's true, that tiny point would be brighter than an equivalent lumen output from a led and would throw further and therefore cut through fog and rain better.

That's more or less the situation with a HID vs a Led.

It is my understanding that with rain and fog, the color temperature of the beam is the biggest factor. Automotive FOG lights are usually amber in color, for a reason. The cool white-to-blue color of most LEDs suffer outdoors, especially in rain/fog.

And, IMO, due to the high CRI of an incand, you see better with less lumens vs an LED.

YMMV.

LEDs emit light on a sort of half-sphere radiation pattern with peak intensity often occurring such that the brightest part of the LEDs emission pattern are not going to be reflected into the central beam. Instead, the brightest part of an LEDs emission pattern is generally emitted as spill beam around the central beam. As a rough estimate on a typical reflector loaded LED, 45-50% of the emitted lumens from the LED come out the front in the form of spill beam, While 30-35% actually make their way into the central beam. The remainder is lost to reflector and lens losses. The differential in intensity between spill beam and central beam is not as dramatic in this example as a typical reflector loaded incandecent would be....

In an incandecent flashlight, the filament emits light in all directions pretty equally, which has a plus and minus effect here. A higher percentage of the light emitted by a filament will be reflected and made part of the central beam. At the same time though, there is more wasted light, since there is often a "hole" in the bottom of the reflector for the bulb to protrude through, light that escapes "down the hole" is basically lost. Seems that this loss is around 15% give or take. The important thing is that the ratio of light in the central beam to the spill beam is often more effective for outdoor work or for penetrating fog or precipitation. To very roughly estimate percentages, I'd say that on a typical reflector loaded incandecent, about 25% of the emitted bulb lumens come out in the form of spill beam, while 40% of the emitted bulb lumens come out in the form of the central beam, the rest is lost to reflector, lens, and "hole in the bottom" losses. When the amount of light in the central beam is ~1.5X the amount of light spread around in the spill beam, the difference in intensity between the two is quite dramatic, the spill beam is relatively soft and non-invasive and non-blinding when it hits closer objects.

It is in my opinion, that because of these differences in final emission patterns; [the ratio of spill to central beam intensity], an incandecent light with a similar reflector size/texture and similar total output is a more effective thrower in actual use than an LED.

Modern single die LEDs tend to have a smaller and more uniform point source to work with than filaments producing equivalent lumens. For this reason, it is not unusual to see central beams on LED flashlights appearing to be more pencil tight. The tightness of the beam will often create the perception that it must be a better thrower than the incandecent, however, when the numbers are run and the lux meters are pulled out we see there is no more useful intensity at range. The incan winds up producing a wider central beam that is just as intense throughout as the pencil thin beam of the LED.

Eric

[edit in] through the use of optical focusing methods, LEDs can have just about any ratio of spill to beam intensity and beam angle desired.

I made a thread about this and I was surprised by the results.

https://www.candlepowerforums.com/posts/3360265

I did a little experiment with the following lights. I have some woods in front of my house but also a streetlight nearby. So there is some healthy ambient light but there is a "tunnel" of cleared vegetation into the woods with a tree about 35 yards in. It is completely shrouded in darkness. This little test gives a good idea how a light can punch through some ambient light to illuminate something at a distance.

You could say this is a throw contest so total lumens don't really matter. I was surprised by the results.

Here are the results in order of results. No beam shots yet, but could possibly do them later.

1. Surefire LX2 - 200 lumens - TIR lens with 200 lumens, this guy puts out a large spot and I can clearly identify foliage details around the tree. Did not have the tint of the P60 but it has so much light it just overpowers everything.

LX2 was head and shoulders ahead of the pack, but the next three were very close in ranking.

2. G2 incan (P60) - 60 lumens - I was really shocked by this one. This light has a fairly tight hotspot that is somewhat oblong. It barely edged out the E2L only because of tint. I kept cycling back between the E2L and the G2 and the moment the incan beam hit the tree and leaves everything was more discernable. It's hard to describe but it's like trying to drive with one eye open and then opening the other eye ... you just have more perception with the incan beam. It edges out the E2L but barely - likely becuase of tint but it made a significant difference in what I could see. I wonder how the P60 assembly in a G2 holds up as a weapon-light?

3. E2L - 60 lumens - Somewhat covered in the above post but this light still amazes me. The TIR really gives this thing some punch and had a bigger and round hotspot compared the G2.

4. G2L + M61 - 260 Lumens - Somewhat surprised at this one, but not shocked. This light is a monster and puts out a wall of light (big, fat hotspot with smooth transition to spill). It definitely wins in the flood category but barely loses out to the E2L. When switching back and forth the E2L has the smallest perceivable edge. The tint on these two lights is nearly identical. I have this light on an AR and the positive side is what it lacks in throw it makes up for in flood - it is no pencil beam! It is more all purpose as it works will indoors and as a wall of light, but does comes in last for throw in this contest.

Possibly put up beamshots soon but thought I would share. I found this very interesting.

Learnings:
- TIR beats reflector for throw (duh)
- Incan beats LED for color reproduction and therefore faster/better object identification ... at comparable lumen output (duh)
- Lights are fun (duh)

Click to expand...

mdocod said:

In an incandecent flashlight, the filament emits light in all directions pretty equally, which has a plus and minus effect here. A higher percentage of the light emitted by a filament will be reflected and made part of the central beam. At the same time though, there is more wasted light, since there is often a "hole" in the bottom of the reflector for the bulb to protrude through, light that escapes "down the hole" is basically lost. Seems that this loss is around 15% give or take.

Click to expand...

I think that the 15% loss estimate due to the reflector opening is high. If you look at an incan assembly, you can see that the support structure for the filament is going to produce some degree of masking of the radiation pattern. In fact, bigChelis has made incan measurements using reflectors from the same source, except one was drilled for a larger opening. Granted, this wasn't as good a test as using the same reflector before drilling and then after drilling. But BigC found no difference in lumens output between a smaller vs larger reflector hole.

It seems you easily can get bigger losses with a poor bezel geometry that masks the beam going out. A shock resistant KT4 TH, for example, loses up to about 20% vs a non shock resistane KT1/2 TH.

Howdy Justin,

Call it 8.7%, not that it changes anything about the larger point being made.

For the example being made here the exact number isn't even close to being critical, it's just being pointed out so that it's understood that there is a loss down there. The point is the dramatic difference in lumen distribution between beam and spill when comparing LEDs that are reflector loaded and incans.

Differential between 65% incan conversion and 80% LED conversion comes from a combination of factors, I contend that the dominating factor is likely to be "down the hole" losses, second in line is probably the fact that more of the emitted light is reflected and takes losses on being reflected, call that the other 6.3%. While bezel geometry would effect both an incan and LED, in fact, an LED will take even more of a hit from bezel geometry since more of the light coming from the unit would be in the spill "zone" and effected by bezel design.

Eric

If an incan really does radiate in an isotropic spherical pattern, then a simple ratio of the hole area vs total reflector surface area would be bounding. The hole area doesn't look like 15% of the total area to me.

The larger point is that BigChelis's IS measurements have not shown any effect of a larger reflector hole size on lumens loss.

Then why do icans consistently loose more in torch lumen conversion than LEDs?

mdocod said:

Then why do icans consistently loose more in torch lumen conversion than LEDs?

Click to expand...

Perhaps a couple testable hypotheses and experiments are in order, in the name of science!

To hypothesize an answer to my own question:

Perhaps it's possible that because so much less of the light emitted by an LED is reflected, but instead, just comes out as spill and only takes the loss of the lens and not the reflector, perhaps that alone makes up the 15% differential.... doesn't seem like that alone would do it, but who knows.

---------------------

One can demonstrate the effects that changes in the angle drawn from the filament to the hole in the bottom of the reflector can have on total output. The difference is actually dramatic enough to be seen the the naked eye. All that is required is an original mag with cam focusing and a dark room with a ceiling.

What if the hole cut in the bottom of a 2" reflector was 1" in diameter instead of just slightly larger than the bulb. The difference between a 5/16" and 3/8" hole might not be enough to show up in testing. Probably within the margin of error from one test to another. Cut a really dramatically large hole and I don't think you'll find anyone who would defend the idea that the large hole wouldn't loose light, but if really large hole looses light, then a small one also can.

There are circumstances where changing the hole size within a particular range is not going to effect total output because the bulb is mounted in such a way that there is a semi-reflective "post" that draws a separate angle wider than the angle drawn to the hole. In this case, there is still a more aggressive loss on whatever post or mounting that the bulb is attached to. The loss is there but does not change as the reflector hole size changes until the hole in the reflector gets large enough to open up a gap in coverage. The reduced reflectivity of the objects directly below most bulbs in reflector loaded applications falls into the category of "out the back" or "down the hole" losses that are not related to reflector and lens losses.

I didn't realize that this was going to get so incredibly off topic. Like we're trying to decode the source of all losses in an incan. It's not even related to the question at hand here. Regardless of where the losses come from, we know that they are higher with incans, and we also know that there is a different distribution of light intensity from beam to spill when comparing LED and incan configuration options.

Eric

mdocod said:

Then why do icans consistently loose more in torch lumen conversion than LEDs?

Click to expand...

Who knows, but that fact alone is not supportive of your reflector hole theory.

mdocod said:

What if the hole cut in the bottom of a 2" reflector was 1" in diameter instead of just slightly larger than the bulb. The difference between a 5/16" and 3/8" hole might not be enough to show up in testing. Probably within the margin of error from one test to another. Cut a really dramatically large hole and I don't think you'll find anyone who would defend the idea that the large hole wouldn't loose light, but if really large hole looses light, then a small one also can.

There are circumstances where changing the hole size within a particular range is not going to effect total output because the bulb is mounted in such a way that there is a semi-reflective "post" that draws a separate angle wider than the angle drawn to the hole. In this case, there is still a more aggressive loss on whatever post or mounting that the bulb is attached to. The loss is there but does not change as the reflector hole size changes until the hole in the reflector gets large enough to open up a gap in coverage. The reduced reflectivity of the objects directly below most bulbs in reflector loaded applications falls into the category of "out the back" or "down the hole" losses that are not related to reflector and lens losses.

Click to expand...

Why would I want to cut a 1" hole in a 2" reflector? Obviously you are going to get "hole loss" in that limit where you start to remove enough reflector. If I removed the entire reflector, I'd probably get a lot of loss. Is this really a surprise discovery?

The situation at hand is with real world equipment with real world sized reflector holes. You have different loss mechanisms in different conditions. You seem to be finally realizing this. I already pointed out previously that the filament mount masks the light in the back direction, obviously when dealing with real world sized reflector holes since I've not been talking about hypothetical 1" holes.

You can debate all you want, but the fact remains that bigC's IS measurements have shown no difference when dealing with real world sized reflector holes, e.g., up to the KD 15mm diam IIRC.

Howdy Justin,

I don't know what all configurations were tested by bigC. He may not have been able to re-create a scenario where the hole seemed to impact output.

There are in fact configurations of bulbs and reflectors that have down-the-hole losses. I personally have a smooth KD reflector from a couple years back floating around here with a poorly designed "flat bottom." The focal point of that reflector is so close to the bottom, that the bulb envelope extends down below the opening in the reflector and light is lost right down around the base of the bulb into the space behind the opening. it's not a ton, but it is happening. The concept of loosing light down there is a real and possible thing even without cutting abnormally large holes, in fact, the hole in that reflector is barely large enough to fit the Big-D bulbs in.

I didn't say that anyone would want to cut a 1 inch hole in a reflector, but if you did, then you could produce measurable losses. The fact that this design variable can create a loss, is proof that the concept is real and not imaginary. The loss is mostly preventable through proper reflector design. If reflector designs that do not have any "down the hole" loss due to proper bulb positioning and a tower assembly are still loosing that extra 10-15% of light, then it's reasonable to assume that it is the obvious differences that are most apt to be the cause. The obvious differences being emission pattern and lesser reflectivity of some of the surfaces behind the bulb.

I guess I need to use a different name for those losses rather than "down the hole," perhaps "poor reflectivity behind the bulb" would be more accurate.
[does this count for digging myself out of this hole?, lol.. Afterall, overall you're right and I'm just desperately trying to save my dignity on this one. In retrospect, in light of what you have said, my 15% down the hole estimate is an incorrect way of stating these losses]

It seems that we agree that the losses are indeed real and coming from somewhere. It would seem that we agree that there is a lack of ideal reflectivity of materials behind the bulb. Whether that lack of reflectivity is related to light going down into a literal hole or simply being absorbed by a ceramic or brass piece is of little consequence in the end, it's lost, may as well have tripped and fallen down a hole eh?

Any thoughts about comparing throw of LEDs vs incans in similar size reflectors?

Eric

mdocod said:

There are in fact configurations of bulbs and reflectors that have down-the-hole losses. I personally have a smooth KD reflector from a couple years back floating around here with a poorly designed "flat bottom." The focal point of that reflector is so close to the bottom, that the bulb envelope extends down below the opening in the reflector and light is lost right down around the base of the bulb into the space behind the opening. it's not a ton, but it is happening. The concept of loosing light down there is a real and possible thing even without cutting abnormally large holes, in fact, the hole in that reflector is barely large enough to fit the Big-D bulbs in.

Click to expand...

Well it seems pretty obvious if you put a bulb partially down the hole you will get hole losses. This is totally different from your original contention that dealt with the fact that incans radiate in a spherical pattern vs roughly 180 degrees for an LED, and thus the back pattern can go down the hole.

mdocod said:

Any thoughts about comparing throw of LEDs vs incans in similar size reflectors?

Click to expand...

Posts #2 and #11 covered the difference in throw -- surface brightness.

Brigadier said:

Automotive FOG lights are usually amber in color, for a reason.

Click to expand...

It's called selective yellow.

There are vast discussions in CPF archives explaining why a incandent light throws better than a LED light. I don't remember the exact threads name but if you do a search you will come up with sensible results.
I'll try to summarise it here.

The size of the source of illumination, which can be made smaller with an incandescent source, has nothing to do with the "throw" when this light is focused into a beam. You can easily compensate the bigger "point source" with a bigger reflector or a dedicated optics.

The reasons for the better throw of an incandescent light are twofold, both related to the spectral composition of the light emitted from the LED source.
The typical white LED is a blue led die covered with yellow phophors. The phospors absorb part of the blue light emitted by the LED die (made with an InGaN semiconductor junction), and re-emit it of a yellow-green colour.
In the LED, the blue and the yellow-green colours "beats" togheter and, in accordance with the Fourier principle, the resulting colours "beats" between them, and so ad libitum.
Theoretically, white light should emerge from the LED.
In reality, the white light emitted from the LED has many spectral "holes", and the various colour components are not evenly spaced along the white spectrum; in any case, there is a blue dominant spectrum line, produced from the initial source (the LED die).

By contrast, the frequency components of the light emitted from an incandescent source are evenly spaced along the white spectrum, and there is no blue dominant frequency - but an infrared one, which is invisible and ininfluent.

As I said before, the reasons for the lack of throw of the LED are two, both tied to the spectral composition of the white LED light.

The first reason is because of the blue dominant line. The light of the sun is white but the sky is blue: this happens because the air of the atmosphere scatters the blue light but not the red light; because the blue light contains - from a quantum point of view - more energy; thus, when a blue quantum enters an atom of air, it re-emerge at far distance, while a red quantum is unschated.
The bluish light emitted from the LED scatters much more easily than the light emitted from an incandescent source of light.

For the second reason of the lack of throw of a LED light, the explanation is evolutionary.
The human eye has light receptors for the three fundamental colors: the red, the green and the blue. The human brain reconstitutes an image based on the information provided from these receptors. These information are compared with already known patterns in the brain; some patterns are "pre-wired", some are learnt from experience.
These patterns are based on the even spectral distribution of the illuminating source, like the Sun or a fire. Evidently our predecessors had sunlight and flaming torches, and a filament bulb resembles accurately the spectral distribution of these natural sources of light.
A white LED light, of the same intensity of an incandescent source, makes difficult for the eye-brain system to discerne the details on an unknown object. This is the additional cause for the apparent lack of throw of a LED light, where the illuminated targed appears "washed out". This effect is more evident outdoors, where the unknown details of an image are more numerous. It is possible to overcome this problem with more LED light power.

Hope this helps

Anthony

I tested the KD V3 15mm vs. a 8.3or something like that MOP reflector perhaps made by Litho.


I re-tested most of my lights over and over and no change in OTF lumens. In both cases I used the same cells topped off, the same hosts, the same UCL lens, and the same test conditions to the "T". OH, I only tested them with the cam on and focused beams.

I thought about it too, and since I only tested with the cam on maybe less light was wasted in both test scenarios. I can try and re-test them with no cam later on, but I have to get another Litho OP/MOP reflector hopefully this week. Right now I am down to just a KD V3 Smooth. There is also a 2in deep fivemega comming next week for more testing.

Good read guys keep it comming.:twothumbs

Justin Case said:

Posts #2 and #11 covered the difference in throw -- surface brightness.

Click to expand...

A typical modern ~200 lumen D26 LED lamp assembly has a narrower beam profile than any incandecent D26 lamp in the same ballpark of output. Assuming both reflectors are textured about the same here. (I have plenty of examples to prove it)

The tighter beam profile doesn't necessarily mean that it throws better in actual use, because the profile of the beam is mostly just a representation of the ratio of the size of the source of light to compared to the size of the reflector. The radiation pattern of the light source comes into play and determines how much of the emitted light becomes a part of that central beam compared to how much of it gets dumped as spill light.

I hold firm on the position that the effectiveness of incans at throwing when compared to similar output LEDs in similar size reflectors is PRIMARILY related to the difference in radiation pattern of the 2 light sources and how that effects their interactions with a reflector.

Does CRI effect throw?
CRI effects perception. It does not effect measurable throw. In actual use, Most are happy to trade upwards of 20-50% of measurable lumens and thow in exchange for better CRI.

Does CCT effect throw?
Only when dealing with atmospheric light obstructing blockages [maybe?]. (fog/dust/smoke). Also, there will be differences in how people perceive and interpret the different CCT, but that does not change the actual measurable throw. [edit in: See post #21 below, this may be totally incorrect]

Does Surface Brightness effect throw? Yes, it's a huge component, but the assumption that incans have a higher surface brightness than modern LEDs may be incorrect.

1mm^2 LED die produces 200-400 lumens with modern bins.

The surface area found on a filament that can compete with this (similar output when driven hard) is about 4-5mm^2

The surface brightness of LEDs surpassed incandescent bulbs several years ago from what this article seems to be suggesting.

However: The LED being a "plane" and the filament being a "cylinder" creates complications for comparing the 2. While the actual flux of the emitting side of the LED may be higher than a measurement of the flux of the filament, the density can not be compared just based on surface area, especially when dealing with how it's going to interact with a reflector. I'm guess that the best way to compare their "density" is to draw the longest line through the light source possible. This would be diagonally across the LED, and maybe down the length of the cylinder that is the filament in a diagonal fasion. The resulting figures will probably far more accurately define their effectiveness as an attempt at being point source.

In the comparison of surface area made above, the same bulb would measure about 2mm, and the LED would measure 1.4mm. That's still fairly dramatic, and means that the LED is packing more light emissions into a smaller space.


Eric