Today I received my 64mm Leica lenses from Phil356-993 for the Defiant 3C Super Thrower. Very nice! I got a measurable 8% increase in OTF lumens.
To demonstrate the difference, I put three common lens types side-by-side and shot this photo of them reflecting light onto a white box. Can you identify them from the reflections?
Left: Uncoated glass
Center: Plastic (stock Defiant 3C)
Right: Leica 64mm coated lens from Phil
This only shows the reflection loss, and doesn't compare transmission loss. Clearly (pun-tastic!) the coated lens wins this one.
The light source is a Defiant 3C with no lens or reflector (mule), so the light intensity is essentially uniform across all three lenses.
Heres a question though, should not the coated and uncoated glass have about the same OTF lumens, anything thats reflected off the inside of the uncoated glass would mostly hit the reflector which would shoot it back out?
Where am i going wrong?
Its not “bad” or “inferior” light that was reflected back, its photonically identical, it was just at an angle that caused reflection rather then transmittance, and should come back to the glass at a different angle then it did initially.
So about 10% was reflected back to the reflector, that should reflect it back out, and 90% of that 10% is 9, so 90 initial plus 9% on second pass should be 99% total.
My thoughts:
The reflector does not reflect 100% of the photons striking it so there are losses there the second time. Some of the light reflected from the lens won’t even hit the reflector itself but things with low reflectivity like the LED and PCB. The light coming back from the lens also is not an an angle close to the notionally ideal point source the reflector collimates. It might have multiple bounces before it even comes back towards the lens and more of it will be at a shallower angle; less will likely get through the lens on subsequent returns. Each subsequent reflection is a chance to be absorbed before it ever gets through the lens.
Photons are finnicky in that each time they contact something, they risk having their energy absorbed as well as reflected. There’s loss in the reflector and the lens (I’m ignoring the small amount of air in the head). Remember, I was only showing the reflection component. I cannot determine an easy way to show the absorption loss on its own.
If the lens passes 90% of the light hitting it, then 10% is what it reflects and absorbs. That percentage will not get out.
If a photon bounces back in, hits the reflector (twice in order to get back out) and gets another chance, then it has a 90% chance of getting out the second time.
Of course you cannot forget that it has to ‘survive’ the reflector itself as well, probably twice, in order to make it to the lens again.
With some assumptions we can work out the odds of a photon getting out of the torch.
- 5% absorption and 5% reflectivity on lens
- 10% absorption and 90% reflectivity on the reflector
- 75% of all first pass photons hit the reflector,
- 25% of all first pass photons go directly from the emitter to the lens
Second pass photons must hit the reflector twice in order to return to the lens
Mathematically the odds of a photon getting out are:
Direct emitter to lens light: 0.25*0.9 = 22.5%
Indirect emitter to reflector to lens light: 0.75*0.9*0.9 = 60.75%
So on the first pass, 83.25% of the light escapes the torch.
For the second pass 5% of the light that made it to the lens was reflected for a second chance:
Direct emitter to lens: 0.25 *0.05 = 1.25%
Indirect emitter to reflector to lens: 0.75*0.9*0.05 = 3.338%
So, 4.625% of the light is reflected for a second chance. This has to hit the reflector twice and pass through the lens to escape:
0.04625**.9**.9*.9 = 3.37%
(I’m only doing one iteration; the third chance reflection amount is negligible)
So, with all of my assumptions, the total amount of light escaping the torch is 86.62% My assumptions were for example only, there’s no need to correct them for the purpose of this exercise. This number is not based on any measurement; it’s purpose was to show that there is very little light getting multiple chances to escape.
Absorption losses in the reflector and lens are probably much larger than the the reflection loss of an uncoated lens.
An absorbed photon doesn’t get a second chance.
While doing this test, I noticed that the reflection amount is less when almost perpendicular to the lenses. With a very low grazing angle, the reflection amount is much higher. This was most noticeable on the uncoated and plastic lens.
Because the reflection was so dim on the coated lens, it was difficult to tell if there was a difference based on angle.
In a torch almost all of the photons hitting the lens are hitting at or very near 90 degrees, this is the best case scenario.
In the photo, the angle is about 60 to 65 degrees, which shows the reflections amounts well.
That is such a neat, simple and effective test and way of pointing out the differences in lenses/optics.
Are you going to do a ‘transmission’ test and how does this differ (angle?) from results you would expect from reflection (which don’t take into consideration the emitter in the reflector), all done with the same light control source?
Transmission test can be done, but I cannot show absorption on its own.
I have an idea of a way to do an effective demonstration of transmission; once I get a chance to do it I’ll post it.
Thanks for posting this. My measurements were basically the same as well and why I started upgrading my lights to AR coated lenses. You may not be able to see a difference or maybe some can with such a small percent increase, but it’s just another step in the right direction towards getting every last lumen OTF of our flashlights. Upgrading crappy wiring to something much better and spring mods also adds to efficiency. All these put together really make a lux meter sing.
You’re welcome! I’d say we think alike. I usually start with the springs and wiring; the low-hanging fruit which is virtually no cost to fix. Reflector alignment and the lens are next.
After receiving my Nitecore EA4W, I realized the difference in coated glass (efficiency and it just looks really cool like it’s not even there).
I’d like to see more coated lens sizes available online (Fasttech, I’m looking at you).
Great discussion here about what happens to the light that reflects back. I think as it reflects and bounces around, the angle to the lens is only going to get worse so you’re losing light not only from repeated reflections but the increased angle will eliminate a lot of the rest of the light.
revive another old, Dead thread… here we go Again….
Kidding
Plastic may be harder to get than glass, but it could be fitted if say a 40mm is located.
I don’t know of any sources but if I find an old junker light with plastic in that size i’ll let you know.
