I watched a flashlight review and at 26:55 into the video he talks about shaping the silicone dome into a cone so the throw gets better. (For a flashlight with reflector)
He says it’s better then a normal dedoom.
Anyone tested this so you can say it’s better then a normal dedoom for a flashlight with a reflector?
Video:
Hmmm, interesting.
Subscribed 
Making a cone on small XPG2 led will be a pita imo.
Oh, it’s David! I like that guy.
Took some screenshots for those who don’t like to push the play button.
I guess the theory for this is that with the cone dome less light will get thrown forward to create spill and more light pushes to the side hitting the reflector and that increases the throw.
This cone dome would not work so great with an aspherical zoomie because on that you would like the emitter to push as much light forward as it can as a normal dedoomed emitter does.
You can increase throw with a waiven collar tho but then it’s always on max zoom if I understand that correctly.
Or I’m thinking wrong here?
You can make it with very sharp blade mounted on lathe.
Looks bad! I guess polishing the dome would help.
Someone needs to try this in a decent sized reflector, back to back with the same LED.
I’m calling BS with the “cone dome”. By his logic, a regular domed emitter would do better than a de-domed, but that is not the case in practice.
I’d be very happy to be proven wrong, & I may just have to try it out myself
After some thinking…looking at the emitter through a conical dome, from a point perpendicular to the slope you get optimum light extraction. There is minimal reflection on the dome-air interface and so the chance of foton extraction is better that with a flat dome.
But that’s just a single point on the emitter. If we look up or down the cone, the light comes angled, so extraction drops. Nevertheless with sane dome angles it’s better than with a flat dome.
If you look at the side though, you get a magnification, thus larger apparent die size.
So I guess there would be larger hotspot with more lumens in it, but nevertheless lower intensity. The spill would be the same or larger; anyway it would be dimmer. The total output would be lower.
But I’d definitely like to see it tested.
I like David Sunshine’s reviews.
Alot of data and information but is he always right?
Like the review about the uTorch UT02 he talks about the choice of emitter at 9:45.
He say that the xhp35 Hi that’s de-domed is the right choice because much light goes to the side and hit the reflector and boost the throw.
And he say if they used the domed xhp35 much of that light would go out the front and flooding out.
But wouldn’t that mean the domed xhp35 would be a better choice in an aspherical zoomie then a xhp35 Hi? (Well any domed Vs de-domed version of that emitter)
Video:
.
I think there might be more light in the hotspot and less light in the spill when combined with a reflector. The luminance porbably doesn’t go up that much (the LED in the video does not look green at all). Higher luminance is needed to actually get more throw.
The dome on a led is effectively a magnification optic that doubles the size of the hotspot. That is why when we dedome it it doubles in intensity or candela.
I think with that much dome left in place it will not match a dedomed led for throw, not to mention the tint shift coming from an angled prism.
Either way still interesting.
You have to take into account, that domed emitter has larger apparent die size - thus lower surface brightness. In case of dedomed emitter, gain from smaller apparent die size (thus bigger surface brightness) is greater than loss of light emitted towards.
No, this is not true. It’s a common misconception.
So what is the truth?…
See here: Flashlight Optics - Dome, Dedoming and Throw
The luminance of a dedomed LED is higher because of photon recycling. The naked phosphor has a different refractive index compared to the dome made of optical silicone. This also pushes the tint more towards the yellow-green part of the spectrum.
Luminance is important because together with the area of the reflector it determines the actual luminous intensity (throw).
The apparent size of the LED (the magnification of the DIE) does influence the actual size of the hotspot though. This is why the hotspot becomes half the size after de-doming.
Would you like to make a few for us to try? If so, how much Premium would you charge? (over the price of the emitter, I mean)
If the apparent size of the LED changes, that means it is also changing the intensity.
If the hotspot became bigger due to the dome, that means the hotspot would also lose intensity because the same amount of lumens are spread over a larger area.
It would go against the laws of physics to have the intensity stay the same while the hotspot gets bigger, unless the dome made the LED produce many more lumens (which tests show is not true)
True.
The ideal situation is when all the light is hitting the reflector and none is escaping trough the front as spill.
True again.
As you said, that’s indeed the function of a wavien collar. To allow (almost) all the light to go forward in a ‘cone’ to hit and enter the lens at its max diameter.
BLF GT? 
I haven’t watched the video so dont know his logic but what you say is (indeed) wrong.
What you want is that all the light is going to the side, and hitting the entire reflector.
A dome or even a domeless emitter cant do that. A cone can work but possible not like this.
If you could make a inverted cone above the emitter, or in the silicon, it could work.
Look at this LED: Witte LED 180 graden 8mm: LEDs-buy.nl het grootste online LED assortiment. It’s a 8mm led that has a viewing angle of 180 degree. It has a inverted cone inside the optic to allow a 180 degree angle. This means that most of the light is going to the side. (and hitting the reflector in our case)
That should increase the lux.
So, what about a cone made with a high reflective material or with a high reflective coating that ‘hangs’ above the emitter and reflects all the light that would go forward as spill on the reflector?
Pls, correct me if i’m wrong. not a optical/led specialist here.
Also found that cree has some interesting documents called ‘design files’ for most if not each emitter. Using this website: http://www.visual-3d.com/tools/PhotometricViewer/ you can view the .ies files.
Sorry, in most situations I would never use my own ideas in practice.