When I had ordered the device, it wasn’t stated the difference between model II and III. This was late November. And in my post #163 it wasn’t stated the difference either; that was January 1st ’22.
You could request a refund on such grounds as the difference wasn’t stated when you made the purchase.
This also applies to Correllux.
It is unfortunate Opple has a poor communication / marketing strategy as the little device appears to work as intended. There’s someone in CS that isn’t at par with effective product description and is playing the consumer. It’s downright deceptive.
From djozzpost #189 I would hope they (Opple) re-issues the device with the newer chipset. By the looks of it, the device sold well and it would be their better strategy to further the development of the Light Master.
Mine also measures 100 CRI on certain 90+ CRI emitters. I have a spectrometer to compare the measurements and the Opple is usually a few percent off, but within margin for error for a general idea of CRI. The x/y coordinates tend to be pretty far off. Duv calculations from these coordinates aren’t very reliable.
If you get Ra100 from LED change distance or restart. Look like it always measuring too high Ra value by 1 to 3. If Ra 95 by datasheet, so Opple shows like 97-98.
There are no rules yet. I found that it takes very little indirect background sunlight to impact readings. Really any other light can change results. People are measuring from different distances. Measuring through tubes or boxes or bouncing light off of ceilings or walls. Potentially even inadvertently bouncing some light off of other objects between the flashlight and the measuring device can change readings. Movement of the device or the flashlight keeps numbers moving. Shift in different parts of the beam can be significant. Other than measuring in total or near total darkness when measuring flashlights there are no rules. There are still variables where you may get a different reading than somebody else. That doesn’t mean that the device is wrong. It doesn’t necessarily mean that you or somebody else is doing something wrong. Comparing this to more expensive or “accurate” devices may not be as straightforward as some assume.
For the most consistent results, measurements should always be done in a dark room away from walls or other sources of reflection and from at least a few feet away. However, even taking such precautions, the Opple’s measurements will always be an approximation. It doesn’t have the ability to read an actual spectrum in 1nm increments like a true spectrometer; it’s just gathering a few sample points and approximating a result.
That said, this device does something that no other device can do at a ~30USD price point. It’s a convenient and useful tool as long as its limitations are understood.
This device was not designed to help NASA study the Sun. Apparently it was designed to help interior lighting “experts” evaluate room lighting or area or light that is cast on to various subjects. Lighting in museums and illuminating artwork is taken very seriously. Having said that I don’t think that holding your flashlight 1 mm away from this thing is the best way to use it.
My comments are based on the Light Master II after some testing on various LEDs (and hot wires).
I suspected the readings were overly biased by Reds. (could be how the RA gets biased higher).
Example #1:
XPL HI 5A3: CCT 3745k, duv calc: 0.0069
Nichia 219c: CCT 3830k, duv calc: –0.0008
These are both in AA Lumintop tools. The XPL appears much pinker (lower duv) to my eyes at all levels. The XPL HI 5A3 is a very nice neutral slightly pink while the 219c is a little yellow/green compared to this light and several others. The XPL HI is known for lack of R9, while the 219c is decent at R9. The x,y coords seem to be overly biased by red content (compared to my eyes).
Example #2:
FWAA shaved LH351D mix of 2x5000k + 1x2700k: CCT 3475k, duv calc: –0.0048
FW3 SST-20 4000k: CCT3667k, duv calc: –0.0047
The 351D is much pinker at the current drive tested. The SST-20 LED is higher CRI and much higher R9.
And then I tested some (3) Surefire Incandescent lights. (stock E2e, 9P driven by 2x18350, CL1499 driven by 2x18500)
This is where it really fell apart.
They are all overdriven (hot wires for you long time flashoholics) and CCT measures between 3450k and 3600k. I can believe this based on eye comparisons to other lights.
But, all three measured x,y coords with a calculated duv of 0.011 to 0.014! They definitely do not appear to have a very positive duv to my eyes.
Is this even possible with an incandescent bulb? This debunks my “overly biased by reds” theory.
Has anyone else tested an incan?
All tests done in a dark room with little stray light (can barely see to type in x,y coords to duv calculator).
All tests done with Light Master against a white wall with white printer paper in front of it to minimize impact of light bouncing off oak desk top.
Distance to sensor is adjusted to get ~15k lux (results don’t vary much by this).
Sigh… not sure this is even worth the <$20 I paid. I know this is not a precision device. I know there is some variation in how/where the measurements are taken.
But, I was hoping to put approximate numbers to what I see when comparing different light sources.
Any idea if the LM II uses the same sensor as the LM III?
Quite right about methodology. Some have been testing in the bathroom. That’s ok for reducing stray light but if you bounce off of the ceiling (assuming this be white, and not a perfect Black Body) there’s also the decor that changes the color or tint shift.
As for the distance, I would think this affects the lux values, which is not what is posted here.
I for one am using my lightbox. It gives consistent values from different readings. However, as the max lux is 50k, there are some differences in CCT and the CRI between the lower mode levels. This is expected as the LED emission is different at current levels. All-in-all, it gives ballpark figures and as mentioned, ± 8%.
Using a light box or integrating device makes sense if you want to see what quality of light you’ll get for illuminating a room via ceiling bounce. For measuring a flashlight, reading from the hotspot makes the most sense since that’s what you’ll be using to illuminate a target. The spill and corona are often very different in CCT and tint from the hotspot so an integrating device will give you an inaccurate result.
Measurement distance is important for the same reason. If you are too close, the corona or spill can affect the reading.
. 
