Yes, mine is 4160CCT at start, also.
.
Remember these specs from emitter manufacturers are for the bare emitter at whichever rated currents are specified in their data sheets. With a bodgy tool like Opple we can only do so much, but if you’re concerned that you’re getting different results, try measuring the bare emitter and see what you get (it would be best to compare now vs. then-when-new of course). If you have a power supply that would be even better but maybe with your driver you can try to pick a mode that approximates the data sheet choice.
Usually won’t see any drastic reduction in output with most emitters unless you’re really pushing them above the rise but that depends on heat dissipation too. Nichias have always been a little more susceptible to it compared to what we’ve seen with most Cree and Luminus but so far the 519A seems pretty robust. Make sure contacts are clean and in good shape, too…doesn’t take much to get enough resistance to affect current delivery a little.
Thanks but all my other LED’s (Getian, Nichia, etc., etc.) measure within reasonable range with the “bodgy” (but “invaluable”) 
Opple. Only 519a is off, interestingly IIRC often on the low side, and I believe this has been observed by others, not just me.
Of note Convoy (and maybe others) sometimes list a range for CCT specs for some of the flashlights/LEDs, eg 4000-4500k, so I understand spec variations. But my Nichia lights (219c, 219b, B35AM) in particular have been close to spec, and I don’t recall ever listed in a range. That’s why I made that remark above. The case of 519a CCT (from Convoy?) is interesting. I actually have brought it up to his attention and even asked if he could personally check the next 519a light I buy from him 
.
Some examples (@Argo note our 519a CCT and Duv match well):
When ranges are listed it’s because that’s the binning that has been purchased (and some are not available in tighter binning at all). Nichia tends to be more on target but honestly I’ve never looked that closely at that part of their data sheets (i.e. I don’t know if they have an “easy white” setup like Cree, etc, but that is a popular and cost-saving option for manufacturers to take advantage of rather than specifying a tighter single bin, especially if they are needing to buy in quantity). But all that is just bare emitter, as sold and usually as advertised…once it becomes a light with a driver and optics it’s all different and many variables come in to play. If you test the finished light and you see changes in output somewhere, it’s hard to narrow it down even if you had an as-new baseline comparison. Coupled with the handy-but-bodgy capability of the Opple sensor and its known shortcomings it’s just hard to chase any accurate conclusions sometimes. Both of my Opples (II and III) were low in CCT and they also varied enough so as not to be considered reliable, but “good enough” in terms of just knowing what range the emitter temp is in. I held off on buying the Hopoocolor until the world settles down a little but that’s something worthy of consideration if you want to get into measurements (sphere, decent meter, power supply…all that can be had for not too much money so it’s the upgrade-from-opple that is the cost outlay and it’s not necessary to go with the expensive photography/cinema models or something lab-worthy). Sorry if any of this sounds repetitious (I’ve probably posted similar before). Not sure if you’re making use of the manufacturer data sheets and grasping the foundations from the bottom up rather than trying to learn all that from the top down, so to speak. It does sound like an investment in a capable measuring tool would be right up your alley, though, and if you ever wanted to do emitter tests it would be awesome for everyone…we seem to have lost most of our traditional reliable testers and their methods/equipment.
I think if you truly want accuracy it would have to be something calibrated against reference, which the Hopoocolor is not. There is really no end to the craziness if you want accuracy. I am just not nutty enough to spend that kind of money.
The value of measuring device could also be in comparison of changes, not just absolute accuracy. But really not the place or the time to discuss this complicated topic or why Opple is more than good enough for me. As we are in BUDGET light forum, not measuring dose of life saving medication, it’s crazy to me to spend more (unless you are in the business for example).
PS Do we need to go through this every time I ask a question on this thread? Could we please just settle on “I’m happy with my Opple since the numbers are accurate enough for me, you are not.”? (At work we would say “could you please not try to help me.” 
)
Well, there’s bodgy, there’s accurate-enough, and there’s reference grade lab equipment. Hopoocolor is pretty darn capable and not terribly expensive and doesn’t have all the cinematic stuff that we don’t need. Didn’t think you were nutty, just that as interested as you seem with the numbers and comparisons that you might want to pursue something that can give them to you better. Flashlights and emitters are in kind of a different place than where most of these photometers and such are, which is why we’ve always found ourselves in the middleground in terms of accuracy, and the Opple certainly didn’t solve that where it matters (duv, mostly). Gen 4 may be a better unit, but we’ll have to see…and if so, whether “better” in this context is really better or not, keeping in mind that their device is really not intended for concentrated beams but rather open air lighting in casual and office lighting environments.
DUV calculation can be done in Excel. The calculation is not complicated and can be done directly in Excel.
Put x value in A2 and y in B2.
Put the following duv formula in C2
=SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)-(-0.00616793ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))^6+0.0893944ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))^5-0.5179722ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))^4+1.5317403ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))^3-2.4243787ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))^2+1.925865ACOS((((4A2)/(-2A2+12B2+3)-0.292))/(SQRT(((4A2)/(-2A2+12B2+3)-0.292)^2+((6B2)/(-2A2+12B2+3)-0.24)^2)))-0.471106)
I know the formula looks ugly and not elegant,but it works well.
Once you verify the calculation, you can copy and paste the formula to next to the columns of cie x and y.
The formula is direct conversion of the code from waveformlighting.com website.
Calculate Duv from CIE 1931 xy coordinates | Waveform Lighting
u = (4x) / (-2x + 12y + 3);
v = (6y) / (-2x + 12y + 3);
k6 =-0.00616793;
k5 =0.0893944;
k4 =-0.5179722;
k3 =1.5317403;
k2 =-2.4243787;
k1 =1.925865;
k0 =-0.471106;
Lfp = Math.sqrt(Math.pow((u - 0.292),2)+Math.pow((v-0.24),2));
a = Math.acos((u-0.292)/Lfp);
Lbb = k6Math.pow(a,6) + k5Math.pow(a,5) + k4Math.pow(a,4) + k3Math.pow(a,3) + k2Math.pow(a,2) + k1a+k0;
Duv = Lfp - Lbb;
Here are my measurements with Opple 3 and calculated DUV using the Excel formula.
!https://i.imgur.com/JvdHonC.png!
:+1:
BTW, I’m using a spreadsheet from TI to draw diagrams:
It’s a bit complicated finding the correct entry cells for CIE coordinates, and the diagram needs some work to make it nicer, but it works well. Maybe there’s better software, don’t know.
Cannga, can do, will do…in fact I’ll just unsub from this thread. By the way, when you constantly go back and edit posts, nobody knows unless they happen to catch it or someone brings it to their attention. You edited pretty heavily on that one. Makes it difficult for forum conversations and future reference. Enjoy your endeavours with opple and learning about lights.
I think I’m gonna wait for the 4th gen device. Micro USB is unacceptable for me in 2022. I sent them the following email:
Add these features for the 4th gen Light Master device:
- Replaceable AA battery (If you don’t offer a replaceable battery, then give us a bigger integrated battery and a USB-C port)
- Duv value visible in the app
- R9 value visible in the app
- Make the sensors more accurate
- Keep the price low (about 30 USD)
I agree and good job; we need more people to spam lol Opple with R9 and Duv result request.
The current battery lasts a long time with this meter, and from their email to me, it seems USB-C is a given.
I have two Opple 3’s, the most important numbers, CCT and Duv, tract well between the 2 Opples during careful comparisons using tripod. The flux (lumen) measurements do not, something that doesn’t surprise me because I have 5 light meters and they read 5 different numbers (2 are used in my Texas Ace Calibrated Lumen Tube and differ by ~7-8%). Fortunately one of Opple out of those 5 light meters give me numbers that match various reviewers’ results for throw calculation, so I don’t feel bad at all owning 2.
For anyone hesitating to buy Opple 3, IMvHO Opple 4 doesn’t really make 3 obsolete, and from that email I have this suspicion that R9 may not be a given. If you feel an irresistable itch for 3, you know what to do 
. It’s invaluable in my flashlight education.
Email from Opple:
Thank you for your support of OPPLE.
- We’ve give this suggest to our Product Manager. (Responding to my request for R9)
- USB-C will be added in 4th Light Master (Responding to my request for USB-C)
Wish you have a good day!
*
Hi. Got my Opple LM Pro (Gen 3) yesterday. I made some measurements on some of my heads / lights. I fixed the light into a fixed vise and position the LM at about 1.1meters away (horizontal).
These are some of the readings that I got.
It would appear that the readings (temperature/CCT) are a bit lower that what is ‘rated’ on each leds… maybe around 150K lower.
Is there a way to calibrate or just accept it and make a computational adjustment with a relative percentage?
I can tell from my experience, the CCT measurements from Opple by itself may not mean much. I came to this conclusion from my own measurements from two Opple units, Gen2 and Gen3. The CCT measurements from the two Opples were often off by more than 300. The measurements from my two Opples are highly correlated (correlation between two measurements were 0.98), so I think it can adequately measure relative differences (higher or lower compared to its own measurements of reference light for example). However, it is poorly calibrated so that absolute measurement value may not mean much.
I measured 24 lights using Gen2 and Gen3 under the same setting. The average difference in CCT measurements is 373 and the standard deviation of the difference is 224. I think the difference is unacceptably high and not very stable. Also, I noticed the difference tend to be higher for higher CCT (correlation between the difference and the average of the two was 0.79).
I have two Opple 3’s and the two track close to each other. For 219b 4500k, one example:
Opple 1: CCT 4511 Duv –0.0086
Opple 2: CCT 4550 Duv –0.0084
I have examples with many other lights that show the same tracking. I will post some more numbers later so you (@akula) could see the range that I see.
There are always some variations from CCT published as manufacturer’s specs. Sometimes vendors would even list a range, for example 5000-5500 CCT, etc. Typically I found the CCT difference to be around 50-100, but sometimes a couple hundred. My 519a 4500k measures 4100k IIRC - wondering at first but found out this discrepancy was also observed by some other owners of Convoy lights with 519a.
I trust my Opple to read within reasonable range, EXCEPT for the flux/output measurements. No surprise as I have 5 total light meters (2 are used with my Texas Ace Lumen Tube) and they read 5 different numbers.
Convoy T3 519A 5000K
Opple G3: CCT 5083, DUV –0.0034, CRI 97.7
Opple G2: CCT 4717, DUV –0.0051
I read the first couple of posts in this thread again and saw the OP noticed 10% difference with calibrated measurements. Off by 373 on average in CCT measurements seems to be in line with OP’s observations.
By the way, some kind of fixture is needed to take stable measurements with G3, as slight movement changes the measurement by a large amount. G2 on the other hand is much more stable.
1 Thank
Agreed. I don’t have G2 so can’t tell, but with Opple G3, I cannot emphasize this enough.
All my measurements have flashlight on tripod, and the Opple receptor lined up so that its receptor is at exact same height as the hotspot and perpendicular to the beam (I imagine this is how factory calibration was done during design phase.). During measurements, I tried my best not to move light and sensor relative to each other.
From your description (more sensitive to positioning), if I have to guess/speculate, I would say G3 is the more reliable. I am not sure the Opple G2’s being “stable” with movement is a good thing. My 2 Opple G3’s read very close to each other in nearly all parameters, except FLUX.
If the sensitivity of G3 was within reasonable range, I would definitely say that G3 is better because it would mean it is catching slight changes in the light. However, G3 numbers jump around too much. For example, if I move my hand in front of the sensor (pointing to the ceiling light and put on my desk) it jumps from 4000K to 7000K before it stabilizes. To me the updates in the measurements in G3 is rather annoyance than a feature. I prefer the G2 to G3 on stability of the measurements.
To sum up, you MUST use fixture for G3 in order to get meaningful measurements, but you can manage getting meaningful numbers with G2 on your hands. (How much difference is there when your hand moves slightly during measurements?)
Another interesting observation is that heat affects Duv and CCT, but not all LED’s to the same degree.
519a 4500k specifically drops its Duv almost 20 points (not too big, but not a small number either) over the 30 minutes that I followed it. Duv –0.0010 cold, –0.0027 hot.