By all means, discuss whatever. It’s all interesting stuff and I don’t have any progress to post on the light at the moment so carry on.
I don’t have much to add on the dome slice approach but I’ve done it in the past and foudn it fairly fiddly. Crucial is polishing the surface and I never really got a good result so I was rather pleased when someone came up with the gasoline/solvent dedome method.
CerealKiller illustrates his method for dedoming an MT-G2 here.
Is it really crucial to polish the sliced dome? when i gas dedome i always slice first with a new Feather Hi-Stainless Platinum Double Edge Razor Blade & the slice look pretty clear after i sliced it, it is the sharpest razor blade i have found yet & the same blades i use in my safety razor.
To my knowledge the “dome” is one material, dome “filler” is a second material, and lastly the glue would be the third material. Typical gasoline de-dome methods actually only remove the bottom glue layer around the ceramic surface; they don’t actually dissolve the filler or dome, but both are softened somewhat by the polymer saturation of the fuel it will absorb. (A two-phase de-dome is capable of dissolving the glue followed by the entire visible dome filler into a liquid.) Anyways, polishing a gel seems pretty awkward sounding if you ask me. lol. But indeed, it might be practical to “clean up” a non-flat dome slice if it happens, since any edge or surface anomaly will carry light forward and out at an odd angle like a bent fiber-optic cable.
Take a gift card or old credit card, drill a hole, ream hole snug on the dome base for the taper, lay it on, then slide razor across dome. Bond wires sit under a mm height, so that much plastic+blade will clear them with a thin layer left. Do not disturb the dome base with excess force! You can create a separation bubble at the die surface, under the gel. It won’t be seen until viewed under magnification, so, light pressure when slicing.
Whenever you do this, set your camera to manual, with lowest ISO you can get by with, and auto-white=OFF. Focus the die image on a piece of cardboard with some grid shown or drawn before and after slice. Be sure to document your tint results, cameras help tremendously (especially if you have an option to save data in “RAW” image format for Adobe software, like Nikon’s do). Some people use white backer boards which is OK, but I find flat-black foam-center craft cardboard stock at WM to be very good for rendering the colors from an LED, for photo-program analysis.
The lens coating company has my sample arriving Thursday through UPS. 
I am going to speak to them about mirror processes later on. They should deal with this sort of thing I would believe.
Oh, LINUSHOFMANN, you can start this test for us with an XHP-70 PLEASE! That should be an easy one. Hehe
I did not seen this thread before. Well done guys :)
I tried shaving and polishing dome with dremmel but not as good as gasoline de doming.
Hi luminarium iaculator good to see you here
Was there any noticeable tint shift when you sliced the dome & used it?
And did you test one of the things we are discussing here, the sliced dome to retain an colder more energetic source for reflective aperture light recycling? I think i remember you experimenting with collars before.
@MEM, i don’t have an host yet so it will take some time before i have an reason to test but, i realized some days ago that i have an colorimeter, the X-Rite ColorMunki Display, that i use for calibrating my monitor/HDTV with dispcalGUI, that is a GUI for Argyll CMS. This should when i figure out what program to use & how, be able to accurately test LED’s for color output & quality EDIT or maybe not.
EDIT
I removed some wrong here information, sorry.
Argyll CMS is command prompt only, if i want to use it for anything other than display calibration and a little daunting to get a handle on.
You could use part of an old projection tv screen. It’s a flat fresnel lens. Cut out a square to set the meter on, with LED aiming at the screen section. I have the screens in my shop if it was actually useful. But i know what you speak of. I had one that hung by the cord over the monitor, triangle shape, and then basically “looked” at different screen sections where the software would place boxes. But a quality camera with images fed into Photoshop should be another crude method to see amounts of colors. A method could be devised to look at the LED light portion only. Then baseline made from starting bin. The harder part to achieve is intensity which is for a lux meter anyway, but color spectrum should be fairly straightforward to plot.
The ColorMunki Display have 2 different modes, one is the Ambient Light Meter mode.
And the spot mode, the normal screen calibration mode when you hang it on the screen you want calibrated.
I don’t have access to any fancy cameras at all, the best i have is an old point & shot camera with very little set up capability. i have tried to take pictures of beamshots with it and utterly failed. So the colorimeter is my best chance.
Your triangular colorimeter is probably a version Datacolour Spyder & those are supported by argyll cms.
EDIT
Hmm i have tested with argyll cms spotread, and read up on other tools, and it looks like most only work best with a proper spetrocolorimeter :(, i had planned to get one to profile my colorimeter sooner or later so i guess this moves it up on the wishlist.
So far i have been able to take lux & colour temp measurement & i get cordinates on the CIE chart.
I got a bit blurred projection because I could not get ultra clear polishing, and maybe small tint shift if I remember right... I did this year ago, and than I toss that in a gass for clear dedome. It won't give better results under aspheric lens for sure.
I am not really sure if polished silicone (XP-G2) can act as more energetic source than clear dome and if it can there will be less output from emitter than with clear de-dome anyway. So if less light comes to wavien there will also be less intense reflection on polished dome.
I am pretty sure that Saabluster(OMG Lumens Deft X) tried that and he decided to leave clear dome on DEFT X.
We should make emitter surface more reflective like thin film of transparent mirror particles spray 
Edit 1000 000 kcd DEFT X here:
Interesting, so you say that the Deft X is with the XP-G2 dome still on?
What does it do with a dedomed XP-G2 then?
And how large is the aspheric in it?, is there only the xp-g2, collar, and a big fat aspheric in the deft x.
And you forgot to remove the k in kcd 
I don’t think so, certainly in that picture it looks like the emitter is dedomed.
I suspect he means “full dedome”, eg. the solvent/gasoline when he says “clear dome”. Is that right luminarium iaculator?
Out of curiosity, what diameter is the aspheric lens used in the deft-x?
I agree that it look like it is dedomed in the picture.
As an extension of the thin dome sliver approach and how it affects the balance between apparent die size and emission characteristics of the light.
I wonder what effect a full dedome followed by a layer of led seal has on the equation.
Essentially you are adding back on a very thin rather flat “dome”, so I would expect that to have some effect. Purely in terms of altering the phosphor layer interface from Phosphor/Air back to Phosphor/Substance.
I realize the refractive index of the led seal would be different to the silicone dome. However it should make a perfect/gapless interface with the die phosphor and by my understanding that should have a somewhat similar effect in terms of emission.
I’ve never applied the sealer myself but maybe someone here could share their experiences with it, does it affect the tint after a dedome at all? Is the Deft-X led sealed?
Your first statement is correct, but knocks the should out of your second statement. If you do not know their reflective indexes are similar, there is no should. Hehe. I have a couple LEDs with seal on them, I do not use it personally. Problem is I don’t know what tint they started at.
Man, Wavien really can’t get their process done right. An optical aperture edge should not look like that, IMO. But hey it worked. Looks like a 20-25mm aperture hole.
I too am curious about the lens. Diameter, and BFL. Or, the clear aperture size that is exposed, even (between bezel edges). The reflective aperture creates a system which is F# limited. What do I mean by that? No matter what lens you use, it is limited by the “F#” created by the reflective aperture. You can change that by opening the hole. As you open the hole, less spill light is reflected back upon the LED, changing the surface intensity. A point exists where optimum reflection exists with any set of a few proper lenses. Of course, closing the hole down to a pinpoint will recycle a lot of light back, but that makes it a high-F#, becoming undesirable/unusable for aspheric lenses with lower F#s.
My wild guess though, is that the lens is around 60mm clear aperture size.
Hi!
I love the aspherical beams ! I have a lot of mod with them, the better is a dedomed xpg2 at 5.5A, wavien collar and 78mm aspheric lens, I am above 1300 kcds the secret is in the focus of the collar, was many hours working on it :_( as they are very very fine adjustments (tenths of millimeters) to achieve the maximum performance…
Yes when I said clean dome(which is wrong expression) I thought on full dedome version, nice and clean with solvent/gass method. We all know that dedomed emitter out throws domed one so no freaking chance he could done that.
Deft X is de-domed like in picture. Lens diameter is 70mm I think.
Probably he did not want to try bigger ones because XP-G2 would have really small die projection...
Do you have this light or you sold it? Please show us this beast because you break record in 78mm size. 1300kcd!
+1
1300kcd is impressive, please do a built thread.
I recently sliced domes on XHP-50’s by using a stainless steel washer that sat on the mcpcb, no pressure on the substrate at all. With a new scalpel blade I found that if I did not slide the blade flat on the washer it worked better… holding it at about a 20º angle on the washer and using a single slice motion left a clean shiny surface. There are micro serrations from the blade edge, but it’s the cleanest I’ve ever sliced a dome. I even reduced height after I figured that out and got neat little 1mm shavings of dome material.
This is a Lucky Sun D80, making 4896 lumens on a pair of Efest Purple 18350’s. :bigsmile:
Hi
Yes, I have in my possession, with the work that led me to make it not going to sell more with the results achieved.
I made long ago, over 2 year inspired by the deft x 
give me a few days to find photos and I build a post…
Sorry for the off topic…
Don’t be sorry we all here want to build crazy throwers.
Fingers crossed you used an reasonably affordable host
I really hope we all can work out the BLF recipe/formula for cheap crazy throwers, better than any manufacturer can make them, just like we have cracked the lumens formula.
We need the equivalent to, high flux bin led, copper DTP mcpcb, fat wire bypass on springs, FET driver or straight DD, fat led wires, high drain battery.
But for lux instead, so far as i see it, high flux bin dedomed XP-G2 in the coldest temp available, biggest fattest aspheric the host can fit in the best quality you can afford possible AR coated (stock or coated after), reflective aperture light recycling, and if you use a very large final lens you can counter the resulting small beam & light loss from led to optic by using several lenses.
The biggest problem with all of these steps is we don’t have good cheap hosts to start from, so very few is experimenting, this is BLF after all