Well…there it is! Happened on a fresh 14500 as well as NiMH. In stepped mode I could only get it to happen on that low 2nd step…everything else was normal. Playing with quick successive power cycles, the behavior varied a lot in both delay and duration…and one time I got the super bright flash. Man that’s an irritating surprise when you’re in a low lumen state of mind! lol. With the light on that second step and then switching to ramping, it did the same thing. Bumped it up the ramp as little/quickly as I could and the next increment behaved the same, the one after that was intermittent. I also noticed a little bit of flutter on the down ramp in that lighting range…don’t see it on the up ramp, though. Weird.
I normally use this light at levels higher than this range-of-quirkiness so I don’t think I’ll bother flashing the firmware although I still want to learn that sometime soon.
Here are some white wall beam shots of SP10 compared with my other lights. White balance was manually set at 5000K. CIE x and y numbers are from Opple Light Master and duv are calculated online
Figure 1. Sofirn SP10 Pro LH351D 5000K vs BLF-348 Nichia 219c 5000K
Figure 2. Sofirn SP10 Pro LH351D 5000K vs Noctigon KR4 Nichia E21A 5000K
The measurements are as follows:
Sofirn SP10 Pro LH351D 5000K
CIE x=0.3546 y=0.3646
duv=0.0028
I prefer lights with CCT of 5000K. This is my first time with LH351D. I am a little disappointed with its tint. Calculated duv appears to be fine, but when I shine the light on my skin, my skin appears to be show too much yellow and green.
I agree, and I think it is because the CRI R9 is not very high…
Excellent compariso photos, and very helpful to have some Opple DUV values for reference.
fwiw, learning to swap LEDs is a game changer. Here is an SP10 Pro album with DUV values: Sofirn SP10 PRO LED swap - Album on Imgur and photos of the modding sequence.
here is a summary of the DUV changes:
.
fwiw, the 519a shows really good skin tones, and the High R9 really makes reds Pop, even when the tint is above the BBL… much better than LH351d, which makes weak reds, and as you noted, poor skin tones.
Thanks for that data… I have 573’s, 503’s and 453’s on hand, does not seem to be a linear change looking at the results of your 573 dropping 1700k and the 453 only dropped 1100k. So, I am guessing that the 503 will drop to slightly below 4000-3800k. I think I will go a de-domed 573.
Thanks for the info. I think the step by step guide will be quite helpful for people with proper skills and equipment. I will definitely investigate it when I am ready for modding.
Green in LH351D doesn’t bother me in high output (more than 200 lumens), but when used in low output I notice it more. This is a problem for me because I use low output most of time.
Among the 5000K LEDs I have, E21A is my favorite so far. It has high R9, therefore reds standout. However, it is a little more than I like. I also own 219B in 4500K, but I think it has a little too much pink. I don’t have proper tools to measure the colors, but when I can definitely notice it when I compare the light with natural sunlight. I usually shine the light on my palm during the day and see the change in tint.
I wonder how 519A compares with my current favorite. I will wait for Hank or Simon to make decent flashlights as I don’t have tools and enough time to invest in modding flashlight at the moment. By the way, I am also interested in trying out B35AM.
A bit of info here, including a really nice chart thefreeman made of his dome-on and dedomed measurements, and a brief followup comment about the fact that the change seems non-linear if you’re looking at color temperature:
I changed out the led in mine for a sst-20 deep red and it works perfectly on AA batteries, wouldn’t try a 14500 though as the forward voltage on the deep red led is too low for direct drive.
The lowest setting on the SP10 pro is actually lower than the lowest setting on the C01R.
The bezel on mine was glued, but wasn’t that hard to break free. Just used rubber wrapped around the light to get a better grip and twisted it off. Nothing like a fenix light I tried to get apart once… the deep red sst-20 reflowed really well on the mcpcb even though the footprint isn’t quite the same. Even centered itself pretty good.