Iâve âcomplainedâ previously that the M21H with 219b 4500k doesnât look as rosy as in S21D. I thought it was because Simon had used a different batch of LEDs for M21H that do not have the same tint. Adding green reflection AR glass helps a little, but the beam tint still is not the same.
I was in for a little surprise when I switched optics in M21H, from the more throwy 12° to a more floody 36°. With the more floody optics, my M21H now looks AND measures identical to S21D.
If I understand correctly (after all these discussions with @QReciprocity42 lol), this is because the tint shift inherent in this dome LED is diffused into the beam with the 36° TIR, instead of being more concentrated (collimated?) into the hotspot with 12° TIR, resulting in a more positive Duv.
Tint and DUV changed by the beam angle for most LEDs, especially for domed ones. This is the famous angular tint shift.
A tighter beam means most of the side emissions gets collimated to the hotspot so it will have lower CCT and higher DUV. Your assumption is correct, as a TIR with larger output angle, especially pebbled surface, tends to mix the light more, the hotspot will contain more center emission which has higher CCT and lower DUV.
3° and 8° TIRs were not good choices for my M21H with 219b x 4. Both have donut holes, more obvious in the 3° TIR. In the 8° donut hole is less visible IRL use, but definitely there with white-walling.
Throw with 8° TIR is indentical to the stock 12°. Probably related to the decreased brightness of the donut hole.
This should hold for well-designed optical systems, but sadly many optics are poorly designed. In the case of Carclo triples, the narrowest (10507) optic has the rosiest tint, while the widest optic (10509) has the greenest tint. There is so much variation in design among optic manufacturers that there is no good rule-of-thumb correlation between beam angle and hotspot tint.
One nice way to tell tint, however, is by looking at the center of the optic and seeing whether there is a convex lens, and how big it is. Generally a larger convex lens gives a rosier hotspot tint. But even this has an exception: the Convoy 5050 20mm optic has no convex lens at all but produces an extremely well-blended beam.
Here are all 6 optics together. Although I ended up staying with the same 12 degree TIR optics that Simon had installed in my flashlight, it was not money wasted. The 24° and 36° have better rosy tint (but worse throw) and should be good for when the rosy itch comes calling.
Throw for 36° TIR is 114m, versus 175m for the 12°.
Iâve also noted the tint change is very obvious in my laptop with OLED screen, but not in another laptop with (lousy) IPS screen.
Iâm curious of the type of driver in the light with 4x219b. Is it the same 6v 5A driver for XHP70.3?
I tried this driver in the S21F host to drive 4x519a in a 2s 2p config and the modes are all messed up. It basically have 4 modes like: high turbo turbo turbo instead of low mid high turbo. I guess itâs because the Vf characteristic of 519a being very different to XHP70.3.
If this light uses the same driver, then either the current sense resistor has been modified, or 219b Vf curve fits it better than 519a.
The Aliexpress web site lists 6V 5A. 6V? Could that be correct for 219b?
Convoy web site doesnât list anything.
My note (didnât write source) says âBuck CC 8Aâ (similar to S21D)
Yes itâs that âfamousâ 219b 4500k. Its output is 1700 lm, similar to all my other lights with quad 219b 4500k like Convoy S21D and Emisar D4V2.
Itâs practical for more closeup work as the throw is âonlyâ ~170m, and the beam is floody, very slightly throwy with the stock 12 degree TIR. It is not going to match the output of any Cree LED. I wouldnât buy a 219b light if output and efficiency are what one is looking for. OTOH, for me, it is the âreference standardâ that I compare all my other lights against.
Itâs a more âmodernâ version of S21D in the sense that it has proper thermal regulation and on-board USB C charging.
I must admit I canât unsee what Iâve seen. The tint of 36 degree TIR is so much better and more rosy than stock 12 TIR for my M21H with 219b. 12 TIR has better throw but not worth it for this low power light anyway.
Iâll replace 12 TIR with 24 TIR next, but wonât be surprised if I eventually will end up with 36 TIR.
For the tint snobs out there, yes no doubt whatsoever 24° TIR optics has a different, more rosy tint than 12 TIR for M21H with Nichia 219bX4.
With 12 TIR, the beam is more yellow creamy, very similar to Nichia B35AM 4500k in M21B. Very nice, but not rosy. With 24 TIR, beam is not as bright centrally (less throw), but now that classic 219b rosiness is back.
If you remember, somewhere here I was complaining that Simon had a different batch of 219b thatâs not as good as S21Dâs. I was wrong, it was all in the optics! Same LED, just different optics causing yellow-green tint shift to appear centrally, making it look more yellow and Duv becoming less negative.
Exactly! Try it in Convoyâs 20mm narrow beaded TIR, or if you have 3 of them, Carclo 10507. Instant game-changer. IMO domed high CRI emitters should never be paired with a reflectorâa well-designed TIR is always better at mixing tint.
This is 219b 4500k and 519a 4500k in S21D both with 60 degree TIR, the most âdiffusedâ and beautiful TIR option for S21D. When I white-wall compare 219b 4500k with another LED, it nearly always makes the other LED appear more green or yellow. The 519a is no exception, but of course this doesnât mean thereâs anything âwrongâ with it, it is just different. Very different and dedoming still doesnât make it as rosy as 219b.
The only two LEDs I have that âsurviveâ this 219b comparison is another Nichia, E21a 4500k, which looks nearly identical but has more tint shift and less output. And lately the âFFL351a 4000k Rosyâ in Firefly E07x and âFFL505a 4000k Rosyâ in E04 Surge.
