I got the light as a gift from contactcr
I sliced that one. I used 3 washers, the final one being a 0.9mm thick 19mm outside diameter very smooth stainless washer I found at Lowes hardware. With Feather Double Edge razor blade. The first two cuts with the thicker washers then it usually takes 1-3 slices with the final washer depending on how thin and even you want it.
I always reflow them on a 20mm star with no solder on the LED pads so it sits flat. If I want to do multiple at once I also have a 33mm 4XP noctigon with no solder on LED pads too.
So after all this testing I think the stand out ones are 2700K, 3500K if you want to keep the dome on.
For slicing I think any of them with Duv of 0.0035 or less would work nicely depending on where you want the tint to end up.
I think I will make a larger order of 3500K and also get the 5000K dog fart from Digi-Key and hope for the best. Still searching for a decent 4K. I have one lead on an unknown bin from China that I may be able to get a sample of.
SPHWHTL3DA0GF4TPQ6 (4000K)
Just some prelim testing w/ Carclo Frosted:
CCT = 4030K (Duv 0.0036)
CCT = 4025K (Duv 0.0033)
CCT = 4041K (Duv 0.0029)
CCT = 4054K (Duv 0.0020)
CCT = 4067K (Duv 0.0011)
shaved LH351D.
Modded Tool AAA:
CCT = 4490K (Duv ā0.0031)
Color Rendering Index (Ra) = 94.2 [ R9 = 69.5 ]
I like this TIR LH351d light,
it has earned a spot on my mouse pad,
The tint is good during the day
and being twice as bright as my sw45k, cant be bad
respect the 351
of note
all the other lights in my edc rotation are 219b
this 351 is the only outsider to make the cut
I particularly like the pebbled tir beam, very even, useful at close range
efficiency test
with a depleted Eneloop AAA at 1.2v in the Tool that contactcr built w LH351s I got 55 lumens on max.
The same battery in my copper tool, powering an sw45k, gave me 25 lumensā¦
so yes, the LH351d Is indeed twice as bright as the sw45kā¦ wowā¦
Looks like MTN has restocked 4000K +90CRI LH351Dās (bare LED). Anyone know if these are the same as the Digikey ones? P6 flux bin, doesnāt say what CCT bin, Iām thinking it might be the same TP quarter bin.
dont know the answer to your questions, but here are some visuals, for reference
the light in the middle thanks to contactcr
max output reference:
the Cu eTool makes 110 lumens with a 4000k 219c or with a 6000 xp-g2, and it is supposed to make 80 lumens w a 219b 4000k
I was checking to see if there was an obvious difference in āefficiencyā as expressed as max lumens on Hi. Albeit the hosts are not identical, they all came suprisingly close to the same outputs, on my light meter, calibrated to an HDS:
73 lm 219b 4.5k Cu Tool w reflector, very pink during the day
72 lm LH351d 5k Aluminum Tool aaa w tir, very neutral during the day, but looks pink compared to optisolis and looks yellow compared to 219bā¦ iow tint is relative
67 lm optisolis 6k JetU w pebble tir, slightly green during the day
Where did you find a tir that fits the tool? I have looked from time to timeā¦
comes on the light from lumintop that way now
https://www.amazon.com/LUMINTOP-Flashlight-Pocket-Sized-Waterproof-Backpacking/dp/B07G98XCG4
they are doing that with the new worm too, now called EDC01
clemence is the master of TiR and cherry picked the one he put on the JetU, I dont know his secrets, maybe ask, or maybe contactcr already can tell youā¦
My tools and rey lights are all reflector lightsā¦
Iām going to plug my own sale thread cause Iām a rebel.
Iām selling a small batch of the 3500K LEDs from this test along with some other spare LEDs from my parts bin:
Digi-Key ādog fartā arrived. Itās almost identical to the Arrow one tested except for about 100K higher CCT ~4800K instead of ~4700K. Duv is otherwise the same.
for a 2700k option I think the SST-20 might have higher R9 than the 2700k LH351dā¦ not sureā¦ I have not compared their specsā¦
here is a link to the Warm SST
Luminus SST-20 J3 JA3 Warm White 2700K CRI95 LED Emitter
pic is a link to the warm LH351d source
How does the LH351D 90cri compare to a nichia 319A (80cri) in terms of efficiency and beam pattern? Thinking about swapping to a higher CRI LED in my Quark Mk IIIā¦
the higher CRI LED will be less āefficientā, inasmuch as it will produce less total lumens
CRI, and especially CRI R9 go up at least as much as lumens go down
imo, it is worth the trade.
now, depending what CCT you are interested in
there may be reasons to choose 219b over LH351d or SST-20
Im not sure yet, which I would prefer, have not tried the SST for example
I do like the Nichia more than the unsliced LH351dā¦
for my pink biases the LH351d needs to be sliced,
that also comes with a lumen penalty
If the die of the 319A is close to that of the 219C, the LH351D will be floodier but probably still have more output/be more efficient even binned at 90CRI. The LH351D has a die size right between the XP-L and XM-L2 IIRC.
Itās actually bigger on the LES vs the XP-L die.
Iāve actually become quite fond of the 4000k sst-20 but I believe those would be even less efficient than the LH351D. I have a few lights with that and while the r9 does lack a bit it seems overall quite nice. Iād really like to keep this as bright as possible and the 319 is doing alright for nowā¦ just love my high cri options!
I think you might have a 319a D340, if thats true, and you go to a sw45k D220ā¦ you would see a 35% lower outputā¦ sounds like you might not like thatā¦ but
let me play CRI advocateā¦
say your light makes 340 lumens now
and with a sw45k, lets say it āonlyā makes 220 lumens
the low CRI is only 54% brighterā¦ that is not really a big visual difference
and you give up all Red renderingā¦ to get those extra 120 lumensā¦ is it worth it?
you know bestā¦ 
lucky for you, that version of the quark is not gluedā¦ tinker on!
Sorry, thatās what I meant unless XM-L2 LES is same size as XP-L? I only assumed larger due to the pad.