Are there some pictures from the types? On Aliexpress there were pictures of the round domeless die version. I bought "Cold white" and "Natural white" option.
I think the main problem is, that no real specs and/or datasheets are known and some LED types simply labeled as "Yinding 5050"...
Cold White ……………………….Neutral White……………………White(warm white)
The Cree Killer 
Glass domed in the middle (it has AR coated glass dome but you can easily de dome for testing purpose if needed with pre tinned soldering iron tip and then put the tip on the top of that glass dome and heat for 30 seconds with circular moves, and after that carefully pry it off with pointy tool like sharpie)
Thank you.
I ordered two of the neutral one and 3 of the cold ones, maybe I really get these LEDs, I don't know :D (Is there any other seller than the one I linked earlier?)
But for now here is the thread about the XM-L2 new design, for these chinese LEDs I'll open another thread with a review. I hope to get my new integrating sphere ready in the next two weeks.
Yes. I agree. We shouldn’t spam this thread. Looking forward for your test! C U 
Nice to see new led tests on BLF, koef3
I have neglected my testing duties in the last two years so I’m glad that you are picking it up again 
If you manage to test the Yinding on the left in the picture that l.i. posted above we can compare our numbers once more because I tested it before. (to be found via my sigline).
Nice to see you here, Djozz :)
I just started with testing of new emitters, after almost five years. A lot happend to me, but now I have better options and equipment to do LED tests. I hope to get the Integrating sphere ready in the next days so I can start testing with proper geometry (after all a box is just not perfectly ideal for measuring light flux).
In the future I want to calibrate the sphere with light sources measured in an official lab to get the best precision possible. Some emitters are yet to be tested...
Kaidomain
I have now dedomed the new XM-L2 more thoroughly and tested whether the dome can be easily removed with a fingernail, similar to the Nichia 519A, which could at least be possible due to the design of the LED.
In short - this is definitely possible, but you have to be careful not to pry off other parts as the silikone dome. Even when cold, the dome (or the rest of it) was easy to remove from the die. Presumably due to the cold condition, a very small portion of the yellow phosphor also tore off, but this does not affect the light image. I suggest this procedure with a warm LED, possibly at around 60-80 deg C.
A broken off piece of the white silicone can be seen in the upper right, which additionally covers the silver-gray substrate around the light area. Interestingly, two layers have been applied: a yellow layer (phosphor?) and then subsequently the white sealant, on top of which the transparent silicone dome is then placed in the further course of production. (It would be easy for Cree to simply bring an XM-L2 HI to market, assuming appropriate binning of the color kits).
The optical properties changed again after the dedoming. At least the light quality in reflectors can be used without problems after dedoming, any rings are usually due to the quality of the reflector. In this example I used a SMO reflector from the Convoy C8.
Noticeable is the lack of shining through silicone around the illuminated area; now the illuminated area is limited to the LED chip only, like with the recent Osram emitters.
The color temperature decreases significantly, as expected with the classic dedoming known from earlier LEDs. From about 6000 K the CCT changed to about 5000 K. Unfortunately, I do not have data from the original LED because the spectrometer had not yet arrived during the original (first) test. However, the light color did not change much when shaved.
The spectral data offer no surprises, they are absolutely typical for an LED in the 70 CRI range, only the red component increases after the dedoming. Due to the color shift into the greenish range, the color rendering indices deteriorate significantly.
After full dedoming the LED turned "green white" and the color rendering deteriorated visibly. This is not nice, but is here due to the color grouping on the part of Cree. By appropriate binning or buying selected color groups, this problem can be avoided as before.
The luminance has increased significantly due to the full dedoming. It is now in the range of an XP-L HI, but the maximum achievable luminance is almost in the range of a Black Flat due to the better overcurrent capabilities. However, this is hardly practical due to the necessary current and electrical power. A worthy replacement for an XP-L HI or a dedicated SST-40 is the XM-L2 in new design all.
Hope this added information is helpful for new modding activities! :)
Greetings, Dominik
Nice… I would certainly try chemical dedoming of this. But I can see from your test that the phosphor layer on the emitter is much softer than on old types of Cree Led emitters. To me speaking honestly - I am really not to delighted with newest OSLON® Black Flat S KW HHL532.TK Jared M test except for very good performance and dirt cheap emitter. This is only single mode emitter imho. I tried to swap new Black Flat on the 2x multimode flashlights. After modding one all modes lost it went to single mode only (although I did insulate it with artic alumina non-conductive glue beneath), ok I said probably I am guilty of that(I thought in my head) then I’ve done second light with even thicker layer of artic alumina and same thing happened? Modes lost on second flashlight. It went to single mode only. It was WTF for moment for me… Still can’t figure that out.
It is not as soft as you might think, with normal handling there are no problems. Of course there is no extra layer of protective glass or film on the LED chip like with domeless emitters like XP-P or Osram ones, but I don't see any problems for using such a dedomed XM-L2 in a flashlight.
The high intensity is very interesting, do you know how many lumens it ends up at, once dedomed?
Hopefully the 5000k version doesn’t end up with such a greenish DUV after dedoming.
Please see the chart in the test. These are the values for shaved dedoming, but even if the light flux is some percent lower for the fully dedomed XM-L2 new (unlikely since the luminance is extremely high at maximum current), this is still very impressive for this LED.
Not just for this led, for every LED. In theory will throw farther than an SFT40 for a given current, with more lumens as well.
Really look forward to putting together a D4Sv2 with these, once a bin that dedomes with decent tint is found.
But it all depends on luminance in the end.
If throw is the only thing what counts, light flux is not the most important thing. (See Black Flat, which does not deliver really high light flux)
Quick question to those who tried both XM-L2 designs - is the tint shift of the new design much worse or similar?
Image from the first post. It looks pretty bad, at least as bad as the first-gen XML2s I’ve had.
I have 1st gen XM-L2 already, I don’t know if the new one is going to be worse or not.
The old (1st gen) XM-L2 have much better light quality (tint shift is much lower).
The new XM-L2 is not recommended if good light quality is crucial. Maybe in warm CCTs the new XM-L2 could be usable.
I was thinking about this one:
But now I don’t think I’m going to risk it since even with the old XM-L2 there’s a visible tint shift.
I heard that LML2.DC has terrible tint shift as well in a reflector, but becomes nice in a TIR.
Maybe the new XM-L2 will behave itself when used with a TIR as well?
Also, do you think there’s a cct that looks good after dedoming?
