I ordered a 219b (listed as “219BT”) triple drop-in, and specified the 4500K 90 CRI variant. What showed up was noticeably ugly. The Opple LM3 says it is ~4200K and 70 CRI with a horrible +0.0076 Duv. I disassembled the drop-in, and the emitters certainly look to me like 219b’s. So now I’m at a bit of a loss. I have reached out to the vendor but haven’t heard back yet.
Any thoughts or opinions? Am I wrong that they look like 219b’s? Nichia doesn’t seem to still host the datasheet to see if there were low-CRI versions available.
Kaidomain is notorious for shipping incorrect LEDs. I’ve personally received the wrong CCT and CRI several times. I don’t believe they have any R9080 or R9050 4500K 219B in stock, since the bare LED listing shows they only have R8000. 219B-V1 was previously available in the sw45 (non-k) chromaticity bin that can have positive Duv. If these aren’t even reaching CRI 80 then you basically got doubly screwed by KD.
For reference, none of the usual sources have 219B sw45k available in R9080 anymore.
Yeah, I awoke too late to the pleasant nature of the 4000K and 4500K 219b, and so when I see somewhere that still has them (or purports to, anyway) I figure I should act fast. That’s unfortunate if they won’t be able to replace it with the correct thing.
If you don’t mind LEDs already installed in quad TIR host, AFAIK both of Convoy’s S21D and M21H have 219b sw45k. The ones in his S21D have Duv around -0.0090 to -0.0100.
S21D runs very hot on Turbo (ok other mode, read caution here), but M21H is ok heat wise. Knowing this wonderful sw45k LED is extinct, I’ve bought 2 of each host, all fantastic.
Low CRI versions of 219B certainly exist! I’ve even come across one in the wild, in some ancient no-name light I saw in some old house. Definitely show your measurements to Kaidomain and ask for a refund.
I wouldn’t be too fussed about the extinction of the 219B sw45k; nowadays many better modern replacements exist, from FFL351A to dedomed 519A. These can handle much more power than the 219B, with higher efficiency, while retaining the pink tint.
Someone looking for a specific bin of 219b is likely particular about this exact Nichia tint, and not efficiency.
While all the LEDs you mention have rosy tint, there is a difference to the rosiness and not two LEDs look the same. For this reason and IMHO it is mistaken to think that a 519a DD 5700k etc. is a direct replacement for 219b 4500k. They are all different, not better, DIFFERENT.
For example and to my eyes, 519DD 5700, has a “redness” tone to it, whereas 219b is truly slightly rosy. This is of course a subjective observation (YMMV), but once seen, cannot be unseen.
Yeah, this is true and something that occurred to me only later. I like the 219b because of the small die and pleasant tint. In reflectors roughly the size of P60 drop-ins, I think it makes for about the best general purpose beam around. I don’t really care if a 519a de-domed can make 600 lumens, 1-200 is plenty in most lights (for me). While the de-domed 519a is still quite good, there are usually some artifacts in a reflector. It is admittedly a very trivial difference, but the 219b is great and why not get it if it’s available. Of course in a triple, this all goes out the window. There’s little to no difference in the beam “cleanliness” with multiple emitters. I didn’t think it through before buying.
Basically I feel I missed the boat a bit on the 219b, so if there’s some inexpensive light w/ them I jump at it. It actually isn’t the pinkish tint of the 4500K I like. I much prefer the 4000K which is negative enough in Duv to look like a very crisp white (to my eye) without looking pink tinted. I can tolerate the 4500K though. So yeah, all in all this was not well thought through. A de-domed 519a 5700K or 5000K (my 2nd fave behind the 4000K 219b) would have been more sensible.
Convoy only lists “219” at this point, and it seems up in the air a bit which you’ll get. In their emitters for sale section, 4500K is unavailable, 4000K actually gets you a 219c. So in lights, I feel it’s a slight gamble. Obviously this KD drop-in was a gamble I lost too, but at the time I didn’t think I was gambling I’ll keep the Convoy in mind though, thanks for the heads-up on it.
(1) Some of us are looking for specific bin of the 219B because they have been recommended this emitter/bin for pink tint, not because they have tried every single pink-tinted emitter and decided this is the one.
(2) If they are looking for this specific tint (in the sense of this particular point on the chromaticity diagram), then up to variance introduced by sample variation and choice of secondary optic and drive current (which is quite a huge variance), other emitters such as the dedomed 519A or FFL/NTG 4200K can achieve the same tint.
Again, two points here to unpack:
(1) What exactly is the difference? If two LEDs with similar chromaticity coordinates differ in the “rosiness”, then it seems that the difference you detect is in the spectrum rather than in chromaticity.
I don’t disagree that there could be a statistically significant difference in spectrum, but it’s good to be precise and specific about what’s different, and in this case, it’s not tint but overall spectrum.
Alternatively, you may be truly seeing a difference in chromaticity, and much of that difference could be just sample variance. That is, you might be picking up on the noise, not the signal. I can see a very visible shift in chromaticity when I simply replace the optic on my S2+, without changing the LED or the driver.
(2) Indeed no two LEDs look exactly the same. Not 2 different samples of the 219B, and not even the same sample of 219B in 2 different secondary optics. Thus, the quest for the quintessential 219B tint is ill-defined.
Do you have a guess of what quantifiable spectral features cause this difference in perception? Is it that the 5700K dedomed has a lower CCT or an oversaturated red region, or that the 219B has a huge blue spike, or something else?
it is good that you are aware of your personal preference.. and I suggest you be gentle with yourself for “missing the boat” on sw45k..
i sometimes try LEDs that I end up not liking.. I consider those mistakes part of my education.. a sort of “seminar fee”, that helps me learn my own preferences.
I hope you continue to enjoy exploring your flashlight options
This is also my take. LEDs are cheap, so it’s nice to try a bunch to explore and learn what one’s preferences are.
Sometimes I even intentionally try LEDs I’m pretty sure I won’t like. I’ve heard for many years about how great the 219B sw45k was, but don’t see anything special about them in all of the color tests I’ve seen. I suspected that I won’t like them due to a higher-than-specified CCT and a huge blue spike (which results in a sterile look and difficulty in distinguishing colors close to red and blue), but still decided to order one to give it a try. I was neither surprised nor disappointed; it came out to be exactly what I thought it would be. Still well worth the money to experience the emitter and to have a piece of history.
YES, you got it. The Nichia 519a 5700k DD has a different CCT and a different Duv from 219b SW45K, by definition the 2 beams have two DIFFERENT colors.
CCT (temperature) and Duv (tint) together determine the “color” of a LED. NTG, FFL, 519a all have different CCT and DuV therefore they land of different spots on the chromaticity chart below, resulting in DIFFERENT colors.
Not arguing one is “better” than the other. Only that they are different, by definition.
Thanks for clarifying. After looking at some color measurements, I am convinced that the 5700K dedomed 519A has a statistically-significantly lower CCT than the 219B sw45k. But I cannot rule out other emitters I proposed as replacements, such as FFL/NTG.
I couldn’t agree with this because the sample variance within the same emitter often overwhelms the true difference in chromaticity. If two 219B samples land on different points on the chromaticity diagram, would you say that the 219B has a different tint from the 219B?
To illustrate with a hypothetical scenario: suppose that 219B samples maintain a tolerance of 4600+/-200K in CCT and -0.005+/-0.002 in duv, while a different emitter (say some NTG/FFL) maintains a tolerance of 4700+/-300K in CCT and -0.006+/-0.004 in duv. If you (as the buyer) have access to the color reading of only a single sample, could you reliably tell which emitter the sample comes from?
The fundamental point here is that claims such as “emitter X and emitter Y have different tints” are actually aggregate, statistical claims, and therefore don’t stand without being supported by a large number of color measurements, large enough to give a reliable estimate of both mean and variance in chromaticity, for both emitters X and Y.
Here’s another perspective: you have samples from 4 different emitters with similar tints, only one of them is 219B. Could you easily/reliably pick it out?
Just curious do you have lights with FFL, NTG, and 219b sw45K that you could shine on a wall next to each other? And take a picture (phone is ok)?
I hope you are not entering this discussion without having done actual experiments? Do you have a 219b sw45k LED in a host?
What are all the LEDs being shown? Best I could tell the 219b in pic above is third one from the left, unless shot taken on colored background and in colored light (not completely dark), pic not locked at 5000k, it’s not sw45k, or the sw45k is in SMO reflector host, etc.
The purpose of experiments is to collect data. If I can collect the colorimetric data from users all over the internet, I’m happy.
The candidates are 219B sw45k, B35AM 4500K, NTG35 4200K, NTG35 5000K, NTG50 5000K, with some being duplicates. I’m curious, what information are you using to arrive at your guess?
For some extra information, here’s a shot featuring the sw45k, alongside copies of B35AM 4500K and NTG35 5000K, not necessarily in that order.
As you remarked, choice of optic can heavily influence the tint. So the notion of the unique 219B tint doesn’t really exist unless you also specify other parameters such as drive current and optic. Otherwise, you get a wide range of chromaticity that samples from many other emitters may fall into.
So was I wrong LOL? Third one from the left first pic. Second one from the left second picture. Beam of sw45k has a specific magenta color that no other LED comes close IME/IMHO unless photographer has deliberately tried to screw test takers over .
Kidding aside, IMHO the final arbiter of beam colors is NOT pictures, but what you actually see. For me, that means with a white wall in complete darkness (removing variables).
Please don’t tell me you don’t have a light with sw45k, or NTG & FFL LEDs to compare head to head on a white wall? To actually SEE? If that’s the case I would suggest we close this discussion.
You consistently picked the same emitter, but it’s not the 219B sw45k. It’s a sample of the NTG35 5000K that has even lower duv.
This seems to suggest that there is nothing mythical/special about the tint (i.e., duv or chromaticity) of the sw45k, and that your characterization of the sw45k was simply “having the pinkest tint possible”.
Photo is from a user who tried to showcase a collection of lights with different LEDs+tints, so I’m inclined to believe that they did not intend to screw anyone over.
There are too many confounding concepts at play. Are we talking about (1) chromaticity (a point on the chromaticity chart), or (2) the spectrum, or (3) perceived color? None of these concepts are equivalent, and only 2 of them are even measurable.
In particular, chromaticity and the spectrum can be fully captured by colorimetric measurements posted on the internet, without the need to see the samples oneself.
If we are talking about chromaticity or spectral measurements as you claimed:
then there is absolutely no need for me to own any of the samples. Owning the samples and seeing their light tells me little about chromaticity unless I also have a meter to measure the samples and get a pair of numerical coordinates.
If you want to talk about perceived color, I am not interested because different people may well perceive the same sample differently, and the same person may perceive the same sample differently under different conditions. In this case, I would be happy to end this discussion, with a suggestion to be precise and specific about the terms you use.
I’ll keep the Convoy in mind though, thanks for the heads-up on it.
.
