Opple Light Master III (G3) discussion thread (Cheap device for measuring Lux, CCT + CRI)

Flashlight Modding and DIY Parts
434

Here’s how I use the Opple LM Pro 3:

  1. Position the Opple on a table and open it with the white diffuser near the top.
  2. Open the Opple Light Master Pro app on my Android phone.
  3. Press Start
  4. Shine a torch at the sensor from 0.5m or 1 metre away.
  5. While the hot spot is focused on the sensor, press Stop in the app.
  6. Take a screenshot.
  7. Repeat step 3 to 6 for each mode, going from lowest to highest.
  8. Go to a PC and type the CCT, CRI, x and y values into a table for each mode.
  9. Use this site to calculate duv based on x and y Calculate Duv from CIE 1931 xy coordinates | Waveform Lighting

Here’s an example of the table:

435

Excellent advice, thank you!

436

Yes. Accuracy is marginal (certainly not enough to discern a few .000 let alone the .0000 people have posted several times). Repeatability it pretty bad but minimizing the drawbacks/standardizing the parameters is the best approach. I have the II, which is less capable with the different sensor but even sitting it on a perch for an open-air reading (as this gadget was intended to be used, rather than concentrated beams) I can scoot it a few millimeters or rotate it slightly on the sensor center point and the readings can change wildly.

Is nobody interested in using cheap styrofoam spheres anymore? Not ideal but it’s been “our” standard for some time and would likely be worlds better for trying to use the Opple for these measurements. Homogenizing the light at least a bit should yield readings that are potentially more accurate and repeatable compared to a concentrated beam in open air where So Much can affect things.

Joann (in the US) still has the 12” ones for pretty cheap. I think $14 last I looked and they’re good enough that they don’t need a ton of smoothing or coping with large pores.

It’ll be interesting to see what they come up with for Gen 4 if that happens. App could use some improvements although I don’t know if DUV is useful for their intended market/product use, esp given its inherent limitations there.

437

Here’s the link for the sphere….$12.99 these days. I think it was Djozz that had several great older threads on how he set up and used a few sizes of these. Still some challenges to understand and overcome just like with a pvc pipe setup, but better than beam-straight-at-a-sensor approach.

438

Agreed. For me, the key for more “sensitive” testing is to NOT hand hold the flashlight, but put it on a tripod.

When I do comparison such as comparing effect of lenses on Duv and CCT (Convoy Green AR vs Purple etc.), I put the flashlight on tripod. Once on tripod, the x,y coordinate readings are stable down to fourth significant figure.

In my experience the consistency of the Light Master Pro is beyond expectation. I guess I could ask for calibration against reference, but that sort of thing is only happening at sensor costing in the thousands.

439

Once the flashlight is on tripod, if I note a change in readings, it’s either thermal regulation or I have touched the light to twist the lens off, etc… Meaning, if you see a lot of fluctuation, it’s because there IS a fluctuation and a good sign of how sensitive the Opple is. If you leave it alone on tripod, some examples of data:

Convoy S21D with Nichia 219b 4500k with Convoy Green AR lens - on tripod and not hand hold:
x3570 y3367 Duv –0.0124
x3570 y3366 Duv –0.0124
x3570 y3365 Duv –0.0125
x3572 y3369 Duv –0.0124

The result I posted in the thread below was confirmed by doing two separate runs, with little change in result as far as differences between lenses. Convoy S21D w/ legendary Nichia 219b :) - Review & comparison w/ Nichia 519a, E21a & other LEDs & lights (D4V2, Convoy S21F). - #296 by cannga . Really at 40 bucks I am not complaining much - a must buy for hobbyists on a budget :innocent: .

440

OK and thanks for the tripod idea for the light. But how do you fix the location of the Opple? As I have found, hand holding it is an exercise in frustration and blows repeatability away completely.

441

You’re welcome - the Opple could just be on any stable surface. I actually was reading the numbers while I took this picture - rock stable CCT and only fourth digit change. Keeping Opple fixed is easy, it’s the light that’s difficult to keep fixed in one spot since you will be doing something with it.

Having said that, this consists of 1 run of a test, the next run it will still be stable but it may read a different number. So personally I will look for a RANGE for anything I measure. And once I’ve done it enough I know this range and don’t lose sleep for variation. (I have a nerdy notebook with pages of data. :person_facepalming: :slight_smile: ). For example the 219b 4500K is around –0.010 +/_ 0.002 or so depending on the light, the GT-FC40 is around –0.006, and so on. It’s a range, but it’s stable and repeatable.

One example: comparing Duv between Convoy Green AR lens and UCL Purple AR lens.

  1. First run Green Duv –0.0120, Purple Duv –0.0100. X,Y coordinates rock stable.
  2. Second run (light and sensor at slightly different positions, on a different day) Green Duv –0.0100, Purple Duv –0.0080. X,Y coordinates different from 1, but rock stable.
    In both cases Purple lens drops Duv by 0.0020. And as I do more comparison, with a different light, 519a 4500k for example, and I notice the drop remains ~0.002. At this amateur level of my hobby, personally that’s all I need to know.

442

Btw there is nothing wrong with hand holding for quick checks. I’m merely pointing out that there is a reason for the fluctuation, and if it bothers you, a tripod will fix it. This is the tripod mount pictured above https://www.aliexpress.com/item/2251832688558005.html?spm=a2g0o.ppclist.product.2.1251M7dGM7dGx6&pdp_npi=2%40dis!USD!US%20%243.63!US%20%243.45!!!!!%402101d1b816617347561721438e003c!65625796532!btf&*t=pvid:0da1446c-94dd-44f0-a906-8874f2bea605&afTraceInfo=32874872757pcpcBridgePPCxxxxxx*\_1661734756

Besides matching position of sensor to light’s hotspot, I’m beyond OCD with respect to having battery full, the light warmed up x amount of time at lower level to prevent stepdown (when comparing lenses’ Duv) or start cold with long wait and recharge battery between runs (throw measurement), etc. So positioning is just one part of a whole tedious but extremely fun process.

Hope this helps and all IMvHO.

443

I don’t see the point of blending because I don’t use any blending when I’m using a flashlight. If you want to measure the spill as opposed to the spot you can do that. I use a small pantry in the house that does not have any windows so I can go in there anytime of day and close the door behind me. I put the opple on the edge of a shelf that’s about hip height. I hold the light with my hand and arm stabilized against my hip to reduce movement. Distance is about 5 ft or 1.5 meters. If I want to test higher light levels that will overpower it at that distance then I have to wait untill after dark.

444

Oli, I guess that’s a good point…maybe more so for CCT than DUV. The aperture on the chip is only 20° I think and the diffuser can only help so much. Integrating the light seems like it should provide better results. Maybe….I mean the sensor itself is pretty slow and coarse so the software only has so much to work with compared to more sophisticated sensors and applies a correction factor that isn’t so granular. This is part of why people can absolutely IGNORE the fourth digit and consider the third one suspect. The datasheets way back in this thread give a lot of info. The newer sensors they have look much, much better…the one in this Opple is long deprecated (I think 2013 was on the datasheet).

People may think I’ve got a beef with this and I don’t really, other than an appreciation for the science and accuracy. The only time it bugs me is when it’s stuck in widely disseminated reviews and presented as accurate information…and actually that’s not the worst thing but I’ve seen people using it to discuss with manufacturers and/or question or complain about lights they have. It’s just not a reliable reference and that’s been pointed out in this thread and elsewhere. Can’t just wish it into being a better device than it is. I don’t know why it’s hard for some to accept that. It’s like the 210E clamp and the “hack” for it…

But to me the coolest thing is that these sensors actually exist at all. They are marvelous little sensors in a very tiny package…pretty amazing really. Some of the other real-world applications are very clever and useful and they may really give a boost to digital photography (phones mostly, it seems).

445

Hello Correllux, no prob, it’s a forum. If you are talking about “reliable reference” then none meets that criteria until you have something like a $5,000 Konika Minolta CL 500 calibrated against reference.

The point is the Opple is accurate enough for one to learn about flashlight behavior. If a reviewer does enough repetition and is careful with testing, sure the info is legit and nothing wrong with it being “widely disseminated.” Same thing with car, audio, photography forum, nearly no one has “reference” material or is a race car driver, but that is IMHO unnecessary for info being shared.

My testing of Duv and throw follows manufacturer’s and reviewers’ data. Nothing is “reference,” yet the information is completely legit for the purpose of this forum IMHO. TS30S throw is 900 range, 219b 4500k’s Duv is –0.0010 range, etc.

The data are repeatable and do not fluctuate much if I fix distance and angle between light and sensor for each “run”, as in my example above. Would you share what kind of testing you’ve done that makes you think info should not be shared?

446

Disagree!

Have you ever used an Opple?

447

I’m a bit sceptical about x,y coordinates, but it’s not about limited precision. Tint mixed colors line up wonderfully in the diagram, e.g.

Rather there’s a bias depending on the true spectrum, but not sure.

448

I was referring more to the standard calibrated light bulb (or in our hobby, Mukka’s calibrated lights served the same purpose). Point being the methods and materials used in lab testing, which we can try to duplicate as best as possible and make adjustments from there as necessary. With Opple being so sensitive and and everyone using different methods (and/or not rigging a system to reduce repeatability errors) results can just be all over the place. With such a narrow angle of entry on the sensor, even with the diffuser, and using concentrated beams in open air (i.e. not an integrating sphere or similar) it really needs some standard of operation, for the ways that we are wishing to use it. We’ve been doing more or less the same with lux meters for a long time now but with those a little error in the setup parameters has a lesser effect on the results even with the cheap devices.

Might disagree there but we’re on the same page I think. “Widely disseminated” is NOT good when the device is not noted for being imprecise and some sort of agreed community use standards are not settled or being implemented. For curiosity, sure, no problem, but we are seeing some manufacturers (some of whom aren’t all that knowledgeable about these finer points of light physics and measurement) seeing out reviewers’ input and using that for designing products, introducing changes, etc. From what I’ve seen some of this isn’t real valid, and I’ll say this with no malice intended but it’s pretty clear that some popular reviewers who have decided to dive into technical aspects of light/lights just don’t have the knowledge to be saying/claiming/sharing what they do. That’s the kind of thing that can take our hobby backwards after a couple decades of working very hard and being blessed with Very Smart people to drive it forward so that we enjoy a lot of the features and materials in lights that we do today. A little device can affect all this? Yes/maybe.

…on the rest, I’d suggest taking some time to read through old tech threads here. Treasure trove of fantastic information and test results on this forum. Between some of those and what is contained in the datasheets for the Opple sensor it’s clear that you cannot expect this device to be accurate with DUV. I think it’s fine for CCT and flicker, it’s good enough for CRI although clearly it’s not accurate in readings or in readings along the temp range. It’s best for open air readings of room lighting, which is exactly what it was intended for. Now…if they adopt that other new 8-channel sensor, that could significantly improve the device (wider angle of entry, narrower half-wavelengths and more channels, plus it can be programmed to suit in terms of what is desired….the current sensor is just coarse and slow and they’re stuck with the factory programming). If Opple put in the effort, the new sensor is far more capable but it will still be doing a lot of back-end correction and assumption because it will still be providing more limited raw data than something more sophisticated. (That last point, simply, is why I say (for DUV) you can ignore that fourth digit and consider the third suspect at times. This is not something you can fix by positioning the light and limiting/removing those other external variables.)

Yes…commented previously in this thread a few times. I now own both II and III but the II is going to be gifted to a photog friend on a budget who just needs CCT. I don’t really see a difference in them in how they perform. I do like having the flicker feature since it’s useful for office measurements (the primary reason I got this in the first place). On the digits, info above in reply to cannga’s comments. Jon, early on in this thread you seemed to accept the findings of others on the limitations of the device, including DUV shortcomings…did you reverse your opinion on that or something? I mean it’s kind of clear and it’s not a knock against the device - it is what it is, and it’s only capable of what it is capable of. The ol’ “lipstick on a pig” saying may apply if we’re all expecting or claiming this device to be something more. Maybe gen 4 will be more capable and more useful for concentrated beams. In the meantime trying to standardize test methods/materials would be best, if everyone wanted to get on board with that.

449

Opple is on sale on Aliex… get one while you can.

I Highly recommend the Opple, and it is on sale for a very good price.

My bet, is once they are gone, there will be no more… do not delay, or you will miss your chance.

The Opple transforms a Wall Of Words about Tint, into numbers that can be compared directly.

Without vague and confusing descriptions like “creamy”, “pure white”, “neutral”, “green”, “not green”, etc.

===
This is how I use it:
The bathroom sink is my “studio”, close the door to make the room dark.

For consistency when comparing two of my lights, I try to get the Opple to read a similar intensity, about 20,000 lux, by either turning up the output, and/or changing how close the light is to the meter.

there is variation in Tint duv, based on whether the light is set to a low output or a high one.

I usually just test at about 20 lumens output… I do not try to test the corona, I just aim the hotspot at the sensor… thats just me… lol

Go crazy… some people use the Opple outdoors, at night, from 4 meters away… others use tripods to mount the light and aim it at the Opple standing on its head… etc… just be consistent.

Dont expect my Opple to match your Opple, as our LEDs are not identical, there will always be variations.

Where the Opple really Shines, is for comparing my own lights to each other.

Its the most fun little tool Ive bought in a long time!

450

Good post Jon. I would also add that for even more OCD people lol, let the light warm up.

My 519a 4500k would drop Duv about 20 points over the 30 minutes after start. –0.0008 to –0.0027. Convoy S21D w/ legendary Nichia 219b :) - Review & comparison w/ Nichia 519a, E21a & other LEDs & lights (D4V2, Convoy S21F). - #362 by cannga
You could sit next to it and watch y coordinate drops (x is mostly CCT change, y is mostly Duv change). Lots of fun. :confounded:

For me, the key is not to hand hold, but set light and device on stable platform and don’t move either light or sensor relative to one another once you’ve started a set of measurement. If both are fixed you literally could see 4th significant figures of x,y coordinates being stable. Like you, the Opple changed the hobby for me, once I’ve learned how to use it “correctly.” Lastly I nearly always run a test twice or more and not stop until I could confirm a pattern. Usual testing common sense.

451

comparing Opple DUV after optics change (watch the dot move relative to the BBL):

Opple data w reflector
.

Opple data w pebbled Tir
.

Cloudy afternoon daylight

comparing LED swaps

.

.

more fun than a barrel of monkeys… LOL!

452

I guessed correctly before looking up the idiom dictionary.
:confounded: :partying_face:

453

Barrel of Monkeys :beer:
https://www.amazon.com/Winning-Moves-Games-1246-Classic/dp/B0B5FBQ7TH/

just about anything is more fun than a barrel of monkeys… :person_facepalming: