2700, 3000 ,and 4000K no problem, 3500K no way.
We are talking here about Nichia and Cree not about some no name. Like mentioned before there is difference in temperature, but opple can not recognize it. It is confusing. So how to check what led seller send for me ?Ahh it is just uselles toy… We get what we paid for…
that makes no difference, LEDs are not precisely identical… three of the examples I posted are Nichias. Their CCT is relatively close, but not identical, and their Tint DUV is not even close
LEDs are not identical to their nominal binning. LEDs are grouped withing a range of variation.
for example, I have seen 5700K 519a that Opple tests say are closer to 5200K. Even when HDS adopted the 5700K 519a, they noticed their batch was testing around 5200K, so they dont call it a 5700K LED, they call it a 5200K LED…
I could go on and on with examples of LEDs that are not testing exactly at the binning they were grouped into
one last example, another Nichia:
in this example the Opple CCT is 8% higher than the nominal 3000K binning.
LEDs that are binned together, are not identical to each other, nor do they all land at exactly the CCT they were binned into, no matter what brand or CCT
2 Thanks
So looks like some sellers dont know what they are selling?
IME with limited number of lights, 519a stands out for testing CCT strangely, very often lower than specs. Precisely the same LED you mentioned, 519a 5700k testing to be 5000k. My other LEDs do have variation, but much less so.
I’ve also observed that, at least in 519a, the CCT changed after use. I say that with confidence because of all the things that LM3 doesn’t do right, CCT has not been much off, at most 200-300 points. It is VERY consistent with testing my 219b’s for example.
Having said that IMHO the variation is not enough to explain @Quadrupel 's Opple not seeing anything around 3500k. That is strange if indeed true. I’m still thinking Opple is right and LED is not 3500, UNLESS he has seen that number 3500 before with a LED, but not now.
Im think we are just fooled by China sellers. And Opple and leds we bought cheaply from China right? ;))
Thanks for contributing more info
part of his testing results may be caused by Low CRI LEDs (he tested an XM-L2)… the Opple has trouble with Low CRI…
Recognized as 86-7 cri
That’s the suffering LM4 owners, us included. 
Kidding aside I have a few low CRI LEDs (Luminus throwers, the Chinese power LEDs, Cree - the usual suspects) and just looking at my Excel spreadsheet now, they test Ok, of course not 100% accurate but no strange behavior. IME, of all the screwy things Opple LM3 could throw at us, CCT is not one of them.
1 Thank
I have a thought. Maybe ask Simon to send you a mixed/blended light like the S21F 519a 5700k/2700k? That would be a great way to check all the CCT readings of your Opple between 5700 and 2700. You don’t even need to worry what true CCT is as CCT blending will go through the whole range point by point.
Just curious, anyone knows if Nichia LEDs are made in China, or Japan? I would guess at least for 219b and 519a, it’s Japan.
yes, we can see .jp in the url:
led-ld.nichia.co.jp
Nichia Address:
491 Oka, Kaminaka-Cho, Anan-Shi, TOKUSHIMA 774-8601, JAPAN
XML2 LEDs can be made in Germany and China
Regardless where LEDs are made, they vary from sample to sample.
1 Thank
I bought that XML2 way back before CREE was sold to China
I got these values from my LM3:
3511K - 519A 3500K in HS10
3593K - 219B 3500K in SP10 V2 (SP10 Pro driver)
Depending on the order code/ bin, XM-L2 3500K probably isn’t strictly 3500K, and then Cree has measuring tolerances on top of that.
The measuring method also affects the result (hotspot, spill, bare emitter), I am not 100% sure but I think the datasheet values is for well integrated light.
2 Thanks
Received Nichias 3500K from Simon and Opple shows correctly. So it means he send me flashlight with wrong tint led before… probably Gearbest send wrong tint xm-l2 back when too … It is good to have Opple
4 Thanks
Has anyone tested this? Got it to work and read Duv continuously with my Lightmaster 3. No more freaking x,y entries for Duv. Will compare numbers with phone App next.
Preliminary testing - numbers not so good so far other than maybe Duv. Just preliminary.
Further testing of Open Light Master web based app. Of the parameters I am looking at: CCT, Duv, Ra, and R9, only CCT and Duv appear to follow the results of the Light Master Pro phone app. And promissingly, it does seem to follow closely, at least with the high CRI Nichia 219b and 519a LEDs that I have tested.
It’s mainly the continuous reading of Duv that I am interested in, so I will continue to test it more. For me, no need to enter X,Y points is a Godsend for comparing Duv at different outputs, AR lenses, etc.
Very simple to use if you would like to try:
XXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXX
1 Thank
Preliminary testing result. CRI numbers are all screwed up. But as mentioned, it’s the continuous Duv I am interested in. So for me still useful because I am sick LOL of manual X/Y entries for Duv measurements.
1 Thank
