I consider it a complement that anyone would think my integrating sphere is a commercial model, but it’s one of several that I’ve made. It is constructed of 2 22” security domes. The sphere is not quite spherical…it’s 21.25” wide, but only 20.5” tall. The inside is coated with 8 coats of barium sulfate/paint. Old lumens has a thread where he is building a similar one. If I were to build another, there are very few things I would do different, so overall, I’m happy with it’s accuracy and functionality. The biggest problem I have (and anyone who has built any kind of integrating sphere/device) is calibration. I’m using ANSI rated flashlights to do a comparative calibration. While this works okay, it doesn’t come close to a using an ANSI rated light source. I test the output on every flashlight I sell, so the performance of the sphere is very important to me…but not important enough to drop $20K on a commercial sphere and light source.
Thanks for the information, you seem to encounter the same limitations as I 
! It sounds like a great device, as good as one can make without spending the multi-kdollar of money :-)
Maybe there is something to that. The domes on the LEDs in newest generation of Cree light bulbs (the ones without a heat sink) have tall square shaped domes.
I wonder what LED this is? The new BR30 flood lights are supposed to use XHP LED.
Awesome stuff! I mean both of your review and the integrating sphere!
I am quite surprised that the tint if the XP-L HI appears to be pure white. I used to shave the dome of an old XM-L (with cool white tint originally) before and the tint started to shift to warmer side even though I only shaved a small portion of it. This is why I am surprised with the cool white tint of this XP-L HI especially its silicon layer appears to be fairly thin. I guess it may have something to do with the die itself.
On the other hand, while some people are happy with the tint of this XP-L HI, I believe there are still certain people who prefer the neutral white tint produced by our ‘conventional’ dedome methods. I know that some of these produce terrible greenish tint, but I’ve also seen some XM-L2 do produce fairly nice neutral white tint too. Maybe CREE will offers different tints for this XP-L HI too in the future, we will wait and see for that.
Beside that, while XP-L HI seems to be quite a good and convenient selection for maximum throw modding project, some XM-L2 reflectors do not fit very well with XP-L/XP-G2 size of emitters. Some examples would be those which require XM-L2 centering piece, like the new version Convoy C8, JAX Z1, HD2010, T08… etc.
these square domes just made me sad as I have accidentally tore the domes of my rgb LEDs in my mouse
why did I even do that…
but these domes seem to scatter light well for a flood compared to sphere which cause some slight pointy artifacts
I had no idea how things work
I think you’re looking at it as if the XPL HI is “de-domed” like we do around here. But, that is not the case. The color binning is done after production. So, the tint will be what they bin it as. It may take a different mix of phosphors to produce the same tint as one with the dome on. But they don’t bin it by the phosphor mix, they bin it by the light output after production.
Yeah I understand that the binning was done based on the light output, that’s why I said it may be due to the die itself - but you phrased it more correctly I think, which is actually the “phosphor mix” in it. 
+1
I don’t think CREE made domed XPL and then dedome them to make XPL HI. That is like you go forward one step and then backward one step.
Making domeless emitters with desired tint is probably even easier than with domed emitters (less factors like focusing of dome…).
Nice tests! It is amazing they can get a 1A out of such a thin dome, and still get the throw. Gotta wonder though if the V2 1A HI is comparable equal to a V6 domed, or better/worse. Always thought this was somewhat possible, just glad CREE did the R&D and accomplished it, though I'm sure it wasn't for our use, more likely for another industry such as auto or spot lighting, etc. Makes sense for those industries because they can reduce the reflector width/depth and still get good distance out of it.
I'd prefer waiting for the V2 5000K (3C or so) to become available - any higher bin than V2 may take a while longer.
Edit: You could de-dome a HI and see what you get 
Pop Tart sprinkles?
Excellent report pflexpro. Thank you for the effort. :)
Nice looking integrating sphere you have there.
If you eat too much of them maybe you will perceive the tint differently!
I think those LEDs in the new Cree bulbs look a lot like a XPL HI with a taller dome.
@pflexpro: thanks for the testing and sharing, really helpfull!
Yes, I would LOVE to see a throw/Kcd comparison test between dedomed XM L2, XP-L and XP-L HI.
I’m planning on swapping the old dedomed XM L2’s in some of my throwers with XP-L HI.
(And I am hoping and dreaming to see the birth of an XP G2 HI…………)
Cheers
Nico
Looks like an XQ-B in the house bulb. Or one of the XB models. Probably the XB-E or XB-G because they’re high voltage at some 23V. 
I have some of the XP-L HI V2 1A coming from Cutter to use in a Quad for the Welight Titanium Ti-Rey or Ti-Ten or whatever he ends up calling it. The Titanium version of an X6. It’ll be interesting to see if throw is improved under the Ledil CUTE-4 optics and what kind of output it makes on an FET +1 driver.
Thanks for this review, very nice indeed.
Quick and dirty test.
I have some XPL HI in 5000K U5-NW but not binned for tint. I did a primary (RGB) spectral analysis of a few of them and found that they would plot somewhere around a 3C and 3D…a little closer to a 3C.
I loaded an xpg2 R5-3C in a host and put the XPL HI 5000k in a host and compared them. Both used an orange peel reflector to blend the light and make tint comparison easier.
The XPG2 R5-3C is in a P60 @ 3.04A and the XPL HI 5000K is in an M1.
I feel very comfortable using this emitter as an XPL Neutral, since it matches the 3C so well.
This photo is Auto white balance:
!
!
This photo is White Balance set at 5000K. This one looks more like what I saw:
!
!
So XPL HI in neutral is a reality today.
Really good comparison testing. Thanks for posting. Excellent setup with that sphere that you built.
I’m upgrading my Cypreus this week from de-domed XP-L to XP-L HI, to fix its tint. Also upgrading the driver from FET to FET+1 to get better low modes. Maybe in a year or two I can upgrade it again to get a higher output bin.
I wonder how much the output will drop, going from a de-domed V6 to a factory-de-domed U5. And I wonder how its lumen output compares to a high-bin domed XP-G2. Think you might be able to test that in the original host?
The old emitters are XP-L V6 2C dedomed, new ones are XP-L HI U5 5000K … and what I’d ideally like is XP-L V6 (or higher) 3D with domes and with large enough optics.
Also, I want your integrating sphere. My current solution is an “integrating milk carton”, which is a bit less than ideal. 
Are these pictures on the side of your milk carton?
No, I’ve got an older model. The missing-person pictures on mine look like this:
