MakeCanon
ModelCanon PowerShot G1 X
Aperturef/4.5
Date & Time2016-09-30 10:26:27
Exposure Time1/10
Focal Length29.3 mm (35 mm equivalent: 104.5 mm)
ISO100
Dimensions2345 x 1565
Size0.8KB
I read thru the Maiden666 review on ForoLinternas as well. Comparing a CW to a NW is difficult - they will not match up well. I'm sure the bins of the XHP70 are different, lower for the NW. The TN32 product page here clearly says the NW is 10% lower, but that's probably just in lumens, - kcd is usually about double the loss of lumens, bringing the kcd spec down to 480 kcd.
So if Dale's TN32 is CW (think so), both Dale's and Maiden666 measured #'s are probably about dead-on matching.
I did choose the CW for absolute power and throw. 
Edit: So yes, if you add 10% to the NW output numbers it would come within about 60 lumens of my measurements. And I did use the premium Samsung 30Q cells, which might account for the rest of the difference. Providing we can assume the testing methods fall pretty close even if quite different by application. I find my P-Trap box to be very close to ANSI numbers of quality lights, this has proven out repeatedly over the past several years. So I trust it. FWIW, I had Chris at flashlightlens.com make me up an UCLp lens to put in my light box, after removing the silicone seal around the heavy plate glass that the brother’s R used when building it. It tested out virtually the same with the high grade UCLp lens. A testimony to the already well thought out build. 
I am back to using the glass plate for it’s durability. The big UCLp lens? Retrofitted to my 15,000 lumen TR-J20.
I agree the numbers make sense: the NW TN42 has about 80-84% of the lumen output and Kcd of the CW version.
But I’m not sure I understand this. If the reflector and LED type are the same, the beam intensity should scale directly with the lumen output.
Pushing light downrange is an exponential factor, it takes 4 times the amount of light to reach twice as far. This is something photographer’s deal with all the time.
That’s really a big head for a single emitter light like this!
So no surprise that it throws better than the K70 although with lower lumens output.
Yes, but I wasn’t talking about the distance. I meant that if you dim a light to half its lumen output, its beam intensity (in cd) will also be half.
Boost or Buck is irrelevant if the output is constant current and the drive circuity is competently designed. As long as the inductor is sized to switch enough current without going into saturation after the battery voltage sags, you’re pretty much set.
IMO, none of these lights should use cells in series if they need unprotected cells. Unprotected cells and series use is a dangerous recipe without a proper battery monitoring system. Of course all of these lights all lack such a system.
I dunno, I just see it empirically (measured) more often than not. Are you sure that's what you measured in lumens vs kcd for a thrower?
Actually it's probably wayyy more complicated than what I said. All sorts of variables - our meters (cheap ones, like under $1000) I believe are calibrated better with CW tints than NW tints, so they tend to read NW's lower and probably in the high range of CW read higher on the meters. There's other factors. Just checked some of my numbers, and sometimes yes, kcd scales well with lumens, but other times I've gotten higher differences on kcd than lumens, percent wise, and I thought as much as 2X.
This is all part of the issues comparing two lights with different tints. Specs from the manufacturers that offer multiple tints are typically not providing the data points per tint, some do, most don't. Chances are they'll quote CW because of the better results.
I can see that stereodude, which would make the Thrunite safer than the Acebeam on the sheer fact that it’s 2 cells in series instead of 4. You can put 2 cells in the carrier and run it, or use a second pair in series for extended run time. Can’t do that with the Acebeam.
I should try other cells and see if I get a noticeable or measureable difference in output, that should tell something about the circuitry, yes?
I can measure a lower mode and extrapolate, but if you have half the lumens then no, you wouldn’t have half the candela. It’s that exponential issue again.
I’ll measure and see what shakes loose.
But, in photography terms… I’m standing on the sidelines at a football game, the punter is directly in front of me in the middle of the field… he’s 25 yards away. Easy flash shot. He punts 50 yds downfield to the sideline I’m standing on, I want to capture the opponent catching the punt, he’s exactly 50 yds from my position. Twice as far away. It’s going to take 4 times the amount of light to reach him for that shot. 4 times at twice the distance.
Candela is a measure of intensity in the center of the beam, so if you cut output in half…. I’ll measure it.
Ok, today for sure this is not my cup of tea. So I’m gonna just tell you guys a few test results and I’m sure some one of you can figure the percentages easily enough…
The TN42, on level 3 of 4 (High) makes 1003.95 lumens and 46750 Candela
The TN42, on level 4 of 4 (Turbo) makes 2459.85 lumens and 702500 Candela
So what does this tell us?
(I have to say here, I’ve tested the candela on this light at least 4 separate times, with the cells at different charge levels two of those times and fresh or almost fresh the other two… every stinking time the meter tells me 2810 with a x100 setting. 28100 at 5M, for 702.5Kcd. Reliable, the most reliable I’ve seen to date)
Everything else being constant doubling the lumens doubles the lux/candela.
Maukka, did you see my test results directly above your post? That doubling concept is obviously not true. I’m pretty sure that double 46.75 isn’t anywhere close to 702.5, even with the differential there.
Dale, I can’t explain your result, but in Maiden666’s review the measured Kcd scales very well with measured lumens and this is generally what I have observed with my lights.
Testing the light’s output with cells at various states of charge would give you how good the circuitry is. Like fully charged, 75, 50, 25%. If they’re all the same the circuitry is very good at running constant current through the LED even as the cell voltage drops and the current draw from the batteries increases.
Think of it this way:
If the intensity of the hotspot doubles when changing modes, the intensity in every part of the beam has to double too since the reflector and emitter are the same. Thus the total output doubles = the lumens double.
Led-Rise’s calculator is misleading. You cannot directly convert lumens to lux. Lumens is a measure of total output, lux is the measure of intensity in the center of the hot spot. A floody 1500 lumen light creates very little lux, whereas a pencil beam thrower making 1500 lumens puts virtually all it’s lumens down the middle, into the hot spot, for far greater Lux.
A simple converter just can’t show that.
Theory versus provenance. The numbers I just got from my meter are truth, without conjecture.
42% of the lumens is making less than 7% of the candela. That’s a simple fact. All the theories in the world can’t disprove it.
Candela, downrange illuminance, isn’t all about the emitter and driver and reflector. It’s about the avaialable light spreading out in a cone as it travels downrange, very quickly losing it’s power.
Science is wonderful when in the lab or sitting at the desk, going out in the field and proving the science is an entirely different matter.
Y’all might be the scientists, I’m the guy out in the field taking pictures of real world results. And the results don’t agree with the science.