Just a shot in the dark here, could physical distance between the emitters also have something to do with providing contrast and depth by shading/defining the edges and textured surfaces of objects when tints vary?
Possible but doubtful as the beams are completely mixed together after a few feet and they are only like 10mm apart at the light itself. Plus I do not really notice the same effect with a single tint in the same setup.
Which is a good point, I think the key to tint mixing is it balances out the good and bad of different tints. For example you get the deep reds of a 3000k emitter and the popping blues of a 5700k.
The only single LED I have seen that can match my tint mixed triple is the 9080 4500k 219B. They are very close with the 219B being slightly rosier. Of course if I was able to get all 90CRI XP-G3’s I think that would balance out some. As it is, the 219B just edges out the XP-G3’s in CRI by my eyes but not enough to outweigh the loss in lumens.
The biggest difference is that maxed out the 219B triple gives me around ~1800 lumens IIRC and the XP-G3 gives me over 3000 lumens with very similar tints.
Check these, might be useful:
Added output test to the original post. This is on the new Virence VR16S1 MCPCB plus an aluminum heatsink w/ fan cooling.
I really envy your new IS Maukka :heart_eyes:
Very very useful for all of us. Since now we have the same LED from the same reel, I can calibrate my output measurements to yours. What’s your MCPCB control temp Maukka?
Just curious how this VR16S1 compared to my old 30mm x 30mm CDTP board?
Thanks,
Clemence
Aahhhhh so that was why I was getting more lumens than I expected on my Haikelite MT01 144AM modded at 1910 lumens, and brightness was even more stable than the XHP50.2.
The forward voltage of 6.14V at 3A is even lower than the XHP50.2’s at 6.22V.
@Clemence, you are a huge help to the community for introducing such excellent LEDs when only OEM used the LED in a diving light, and everybody else doesn’t.
I’ll probably be buying a bunch more of your LEDs and MCPCBs when your 20mm MCPCBs will come.
20mm? Did I spill the news? 
They will come… eventually.
I think. Perhaps? Hopefully.
I don’t think the MCPCB really heats up much above room temp (26-27°C) at the binning point (1400mA) during those 70 seconds it takes to reach it in the test sequence. Will have to verify with thermal camera. Haven’t checked with the old board.
It under performed 23 lm from the rated 1000lm/1,4A. So far only the CRI exceeds the minimum rating.
The 2000K E21A also 3 lm less than specced. Nichia binned it using 10ms pulsed current though
The best I got was 998 lumens at 1400mA when taking the measurement immediately after turning the power on. It still takes a couple of seconds to get the reading.
The Samsungs and Luxeons seem to be above their minimum rated flux though even after compensating for the temperature difference. They bin at 85°C. I don’t like calculating the compensation from Lumileds’s temperature vs. flux graphs with Photoshop ruler, anyone know if they are available in table format anywhere?
Well, I’m sure it’s up to spec with 10ms pulsed current. The newer batch of 144AM are 2% more efficient (1050lm vs 1071lm at 1400mA).
Another method to find point(s) from a graphical curve is this:
Got this link from Texas_Ace. Very useful when dealing with secretive manufacturer like Nichia
- Clemence
Oh wow, that is a lifesaver, thanks!
Indeed, it is amazing. I wish I could remember who sent it to me, it was someone else on here during the GT design. Might of been 5AR.
Great tool, thanks!
Superb with interpolation from charts by secretive reviewers who don’t publish raw numbers. 
Today I asked for E1000 or E1100 flux bin for sm403 and sm453 144AM/144AR instead the E900 like what I got earlier. And Nichia simply said they can’t give it to me for my order quantity (3000 pcs). The yield is too small for such high flux bin.
When I said I need the highest bin possible to compete with existing XHP50 I received a short answer: No it’s superior to XHP50!
Then I sent them the output vs current graph of 144AM vs XHP50 from Djozz’s test which indeed showed that 144AM was superior to XHP50 below 4A. I received another answer: Sorry we can’t accept your evidence since it’s far beyond datasheet rating. It’s better than XHP50 (XHP50 CRI 90 vs 144A R9050).
Enough said, can’t argue with them…. 
- Clemence
Can they give you E1100 in smaller quantities?
No. It’s random between E900, E1000, and E1100
- Clemence
Reviving an old thread.
Can anybody recommend a good flashlight to be used as a host for a Nichia 144am?
Open to 18650 and 26650 single cell lights, on the compact side with a good efficient boost driver.
Some lights I’m considering are the Sofirn SP33, or Thrunite TC20.
As the Nichia 144am is prone to yellow donut in the middle of the beam, wanted to check in and see if anybody has successfully modded a light using this LED in a reflector without the need for DC Fix.
The only reflector I’ve found effective for perfectly smoothing out the egg yolk hotspot has been the one from the Nitecore TM03. The one COB LED optics Clemence sent me for E21A tint mix testing was also superb for it.