I had that Cree LED, but second version XHP 70.2 ,and in “Neutral White” CCT. That particular LED was the worst green hell in my light collection. It was going back to Amazon even if the efficiency were quadrupled . Kidding aside, IMHO low efficiency is part and parcel of high CRI with great tint LED (that article you quoted even mentioned it). I am not surprised that any other LED have higher efficiency than GT-FC40 or Nichia; in other words, good to know, but personally, low on the “priority” list.
It seems they have fixed the tint shift for Cree 70.3? For fun, please post pics of beam shots of this 70.3 next to your GT-FC40 and Nichia 219b. I don’t care too much for CCT 5400 (too flat and “sterile”) but would be interesting nevertheless.
BTW with 519a and GT-FC40, preliminary trials with my Opple appear to show Duv MORE negative and Ra ~ 4 points higher than what Simon measured (219b results are closer). Whether this is due to LED variation, measurement difference, or systemic error of the device, I don’t know. But personally at this point, other than CCT, I still take other Opple results with a small grain of salt. I trust photography comparison more, for example.
Yeah,Opple is Opple. It isn’t lab tool for sure. I’ve bought about 10 different bins of these 70.3 Hi’s for testing. This 2D 90cri is the best one by far. These are domeless, so no tint shift.
Hardest part is to find a bin, that has them all. Tint, cri and effiency. I would have preferred this to be little warmer but this tint is so clean white, that I don’t care.
With GT-FC40’s it’s also lottery if you get a good one. I have bought 5 4500K versions at three separate occasions and first 2 of them were good really good, rest were green. I also have 2 3000K versions. Both green. And sustained output and throw of these wasn’t enough for my use case.
I will soon receive few more of these. Hopefully they still are the same bin, because they sold out fast last time and I only bought one for testing purposes.
I took beam shots of the 2 lights to show the different beam angles. S21D almost 180 degree spread. Still I am a little surprised the difference in throw and overall brightness perception.
The S21D in this picture has the more throwy Sofirn IF25a lens btw.
1. I do find its reading of CCT, brightness, and Ra very helpful (if possibly a little optimistic with the Ra sometimes). It’s the Duv that I am a little cautious about and I still think back to back “real life” comparison or photography are more “reliable” ways to compare LED colors (not Opple Duv reading).
2. Wow - did you buy LED, or are these actually 5 separate Convoy lights :+1: ? If you buy LED’s, it’s understandable that the bins might be different, as with Cree or any other brand. That said my 4 Convoy GT-FC40 lights have beams that are instantly recognizable as belonging to same family. Even the Duv readings (posted above), if possibly not accurate, are at least consistently negative.
I always knew from real-life use that the GT-FC40 has a purple/magenta tint, but have not been able to capture this color on its beam shots (they always look more brownish - see picture above). This has been a puzzle to me, that photography doesn’t support what I see.
But I’ve finally cracked the code, by way way UNDER-EXPOSING the central area of the beam, which usually is very bright.
Below are beam shots of my two M21F’s, one with the very negative (Nichia 219b-range) Duv of –0.015, and the other a more moderate Duv of 0.0050. The magenta/purple tint is clearly captured. Now there is no doubt; finally I’ve exposed the Nichia imitator lol. Equally great that the M21F light with the more negative Duv is actually more magenta ie photography correlates with Opple’s measurements.
Still burns me how CCT can vary so much with a given emitter, by production batch. Manufacturers really need to aim for greater consistency… but, I guess it all comes down to tolerance & demand. There aren’t nearly enough people OCD on tint.
I did not record the x,y coordinates the first time (prob too excited with the then new arrival Opple Light Master Pro/Series III) so re-ran the test a few weeks later. Yes you read correcly somehow I did end up with four GT-FC40 lights :person_facepalming: lol.
Continued to be impressed with the Opple’s consistency. An absolute must have for enthusiasts IMHO - who don’t want to spend more than 40 bucks that is. Make sure to click on the picture that says Pro (3) Page Not Found - Aliexpress.com
I have the one GT-FC40 outlier that has very negative Duv of –0.0148, in the range of Nichia 219b 4500K (you could even see it in the beamshot above Click HERE ). Overall I would expect the Duv to be less negative, in the range of –0.0060 as seen in the other 3 GT-FC40’s that I have. CCT of all 4 is in the ideal range for an all-around EDC with soothing yet clear beam color - 4500K.
Second Run:
M21B
CCT 4579 Ra 98.7
x 0.3560 y 0.3466 –0.0068
M21F
CCT 4370 Ra 97.9
x 0.3595 y 0.3339 Duv –0.0148 (!!)
M21F (second light)
CCT 4481 Ra 98.4
x 0.3591 y 0.3491 Duv –0.0066
M21E
CCT 4395 Ra 98.2
x 0.3623 y 0.3520 Duv –0.0063
If you own any of these Convoy lights and the lens does NOT reflect green (apparently earlier ones had purple reflection), check out this comparison between the Convoy AR lens w/ green reflection and a UCL lens with purple reflection +*HERE and white-wall beamshot HERE .
Summer is gone so it’s time we take the flashlights out for our evening walk. I know I sound like a Convoy salesman at times (or always lol), and sorry but I can’t help it! . Many times it feels like Simon knows what I want and make a light for me.
At any rate the M21E with GT-FC40 has everything that I could ask for: beautiful color rendition, warm beam color, perfect throw range. It is bright but not too blinding to people in opposite direction (they hate my Nitecore with SST40 6000k). If you are new to flashlights this should be the first light to buy IMHO. A true perfect all around EDC. If you don’t believe me please just google “GT-FC40 review” and read the raves.
Most people would probably prefer M21F (less cumbersome, prettier) but because my older eyes like everything bright, I have a slight preference for E. Also IMHO warmer CCT LED like this 4500k needs to be a little brighter than 6000k LED for better scene “clarity”.
Here is the light next to a 60 year old watch I inherited from my father. Due to life circumstance (war & immigration) it sat un-used for 30-40 years. I sent it to Omega service to restore and they asked me what happened, did owner wear this watch lol. They did a wonderful job, down to original Omega leather band. Amazing that this old watch is still off by only a few seconds per day.
You might have heard of the famous Convoy’s sustained brightness. Here is another example of the algorithm. Seems like thermal regulation kicks in at 55 C, but power is decreased to a level that is fixed, and does not go lower regardless of temp. As result the temp could sometimes go above 55. I recorded around 60 C at the head for many of my Convoys.
Of the lights that I have, final brightness level is usually about 1/3 of max brightness, using ceiling bounce measurement. The H2 headlamp with Nichia 219b is an exception, where it is 60% of max. For ease of comparison, max starting brightness is assumed to be 2000 lm for both lights.
Because I personally like brightness in use, this algorithm is actually one of the reasons that I like Convoy so much. Note also the stepdown of M21E with GT-FC40 is more abrupt compared to S21D with Nichia.
That’s a classic. One of my all time favorite movies.
More flashlight fun pics. I do wish for more variety of finish for these bigger Convoys. The Olight 3S’s Green is beautiful to me and am asking Simon if it’s possible.
With the recent addition to my flashlight emergency fund , I am thinking about the 5000K version of GT-FC40.
I’ve added M21F runtime to the above curve. The difference between M21F and M21E as far as runtime is really not much. From 100% start, M21F steps down a minute earlier, but settles to the same excellent Convoy sustained brightness, around 35% of max.
The main difference between 21F and 21E therefore is not really runtime, but the beam character. No surprise here, hotspot vs spill: M21E with a bigger reflector has brighter hotspot (more throw), and less spill (less flood). M21F has dimmer hotspot, but brighter spill, resulting in a more even field of vision, characteristic of a more floody light. The difference is obvious in back to back comparison. For medium and close up around the house work, M21F is better IMHO.
Both M21F and E are top notch medium throwers with high CRI and above all a tint that is closest to the famous Nichia 219b rosy tint that I’ve come across. Although I have other Nichia lights, and throwers, and of course the “ridiculous” TS30S with SBT90.2, M21F and E are the lights I use the most. I do have a slight preference for M21E because it is so rare to find a thrower with such fantastic beam color, and I like brightness for my walk. But that’s just me. Up close the M21E with its bright hotspot could be a little blinding.
Both a must have for any light collection IMHO; you can’t go wrong with either.
If that was already there and I missed it, my bad.
However, I don’t use 100% on either version of the S21D because it gets too hot to hold, so the comparison I’m really after is the 35% mode of each (aka the real-world useful setting.) Just curious how they stack up.
The output on the 12A looks like it stabilizes at around 300, wondering if the 8A is identical or a bit brighter (because it doesn’t waste all that power in the first two minutes…or does it?)
. Kidding aside, IMHO low efficiency is part and parcel of high CRI with great tint LED (that article you quoted even mentioned it). I am not surprised that any other LED have higher efficiency than GT-FC40 or Nichia; in other words, good to know, but personally, low on the “priority” list.
. Many times it feels like Simon knows what I want and make a light for me.
