what was the answer from Ali?
DUV measurements seems closer to reality for LM4. LM3 consistently measure lower DUV values.
The modulation seems insignificant to measure flicker. You should use a PWM light for a better comparison.
I mean, for DC the frequency reported is pretty much irrelevant. The app is basically deriving frequency from the noise measurement. What they could do is add heuristics and report 0 back at you, but I bet you’d be angry at that too…
Some people just can’t be pleased. It really sounds like you should buy the $1600 meter and get it over with. (Oh wait, that meter can’t measure flicker, LOL)
Instead of trying to get a free product from the seller and raising the price on the rest of us. Just return the product back to the seller and get your refund.
2 Thanks
Try testing an incandescent?
My incandescent maglite measures RA100 and R9 100.
I have tested several lights and I think I found out some interesting patterns of R9 on LM4.
- I get reasonable readings of R9 on low output. (My LH351D measures R9 of 70 in low output, which is reasonable)
- On turbo output, all lights(both low and high CRI) displayed R9 of 100. So, R9 behaves weirdly if the led is driven in a certain way. Weird measurements usually happen in higher output.
1 Thank
2 Thanks
I need to sort through the screenshots but I took some reading this weekend, and this was my experience too. I find that trying to get close to 20,000 lux on the meter seems to give readings that are accurate to known specs.
I just got my Opple 4, there are definitely issues with it. DUV on the Opple 4 seems to be higher than it is supposed to, on most of the LEDs I tested other than the SBT90.2 and SST20 5000k, as well as R9 readings that are low when compared to the specified values of Nichia LEDs. All the LEDs in the test originally came from Hank, so I trust them being the claimed spec.
I can take requests as I might have a light with that LED already (osram w2, sft40, ts10 csp2323, 219c, sfn60, lh351d, etc), but just couldn’t try it today for time and (opple 3) battery constraints.
All measurements taken at the same ramp level for both meters, aiming for the magic number of 20k lux, which seems to be where these devices work best. Text in bold are especially unusual readings with a comment below the table.
| LED | Opple 3 | Opple 4 |
|---|---|---|
| 519a sm573 dedomed (R9080) | 19865 lux, 3951k, Ra 97.4, DUV -0.0051 | 19755 lux, 3940k, Ra 95.7, R9 72.2, DUV -0.0027 |
| 219bt-v1 sw45k (R9080) | 19720 lux, 4602k, Ra 96.3, DUV -0.0087 | 20584 lux, 4655k, Ra 95.2, R9 73.2, DUV -0.0052 |
| B35AM 3500k (R9080) | 20198 lux, 3172k, Ra 96.9, DUV -0.0040 | 19606 lux, 3233k, Ra 96.2, R9 78.7, DUV 0.0014 |
| E21A 2000k (R9080) | 12971 lux, 1960k, Ra 97.2, DUV 0.0016 | 20033 lux, 1939k, Ra 89.9, R9 60.2, DUV 0.0066 |
| E21A 2000k+5000k mix | 19656 lux, 3523k, Ra 96.3, DUV -0.0114 | 19942 lux, 3609k, Ra 95.9, R9 72.6, DUV -0.0061 |
| E21A 5000k (R9080) | 20034 lux, 5199k, Ra 98.2, DUV -0.0039 | 20251 lux, 5159k, Ra 95.2, R9 88.1, DUV 0.0008 |
| SST20 4000k (95 cri) | 20131 lux, 3742k, Ra 97.1, DUV 0.0018 | 19676 lux, 3816k, Ra 96.1, R9 89.8, DUV 0.0044 |
| SST20 5000K | 20016 lux, 4505k, 57.3 Ra, DUV 0.0169 | 20315 lux, 4384k, Ra 65.1, R9 –, DUV 0.0149 |
| SBT90.2 | 20963 lux, 5009k, Ra 60.6, DUV 0.0138 | 20112 lux, 4995k, Ra 65.9, R9 –, DUV 0.0118 |
| XHP70.3 HI 4000k (90cri) | 19767 lux, 3981k, 92.9 Ra, DUV 0.0050 | 19868 lux, 3898k, Ra 93.5, R9 79.2, DUV 0.0065 |
| Wolf Eyes Eagle-2 (incandescent) | 20786 lux, 3559k, Ra 98.5, DUV -0.0012 | 20183 lux, 4411k, Ra 100, R9 100, DUV -0.0118 |
- The SST20 4000k are not actually that green, barely above BBL at mid ramp, exactly on it in turbo.
- Opple 4 did not show an R9 for the SBT90 or 5000k SST20, no matter how much I tried, maybe they have patched it to not display R9 if Ra is below 70.
- The B35AM 3500k is clearly below-BBL, using my eyes.
- Sensor on the Opple 3 maxed out very easily with the 2000k E21A, this was the highest-lux reading I could get with it.
- I have no idea why the Opple 4 read 4400k cct and a ridiculously negative DUV with the incandescent, the coordinates are x=0.3594 y=3395. Maybe it wasn’t programmed with data for high-cct incandescents.
Update 2023-05-09: Going to try and return it, hopefully a full refund, and re-purchase once the callibration issues have been worked out (might be another version with a better sensor).
5 Thanks
In my experience, a lot of measurements change depending on the output levels of a LED. If you are getting readings that are similar to known specs at certain lux levels, it is probably because LEDs output levels are near the standard levels at which the specs are labeled.
However, considering the range in which Opple is configured, I believe it works best in the lower range of lux levels, below 20,000 lux. Opple is marketed for measuring indoor lighting, and typical indoor lux levels range from 400 to 1000, with bright indoor levels ranging from 1000 to 5000. Therefore, I think the most accurate measurements can be obtained within these ranges.
If my assumption is correct, the assumed reliable range is certainly a limitation for measuring flashlights, because lux levels taken at close range can easily exceed 10,000 lux. To measure flashlights, you may have to put the flashlight further away from the sensor and control the ambient lighting, or use lower output levels of the LEDs. However, as noted above, low output levels can produce measurements that are quite different from the specification.
2 Thanks
Welp, i sent my LM4 back to AE. They sent me a msg that they will make a new version with a better chip…Enough for a refund…
8 Thanks
Excellent, thank you for letting us know!
2 Thanks
Not a great start for a LM4…but I hope they make you happy in near future
Thanks for sharing. Not good news if true. This is different from what being stated on Alix, which seems to be an emphasis on correction with software update?
Is software correction not mentioned at all on their message to you?
"…We have developed a new app for Light Master 4 called “OPPLE Smart” not “Light Master Pro”, and we have to add data and test code of “OPPLE Smart” let it can read more measurements of different light…
It takes time, estimated app updated on 30th, May to fix some bugs we found, then a small iteration optimizations updating will be done in the following 30th June, 30th July. (It can only be updated once a month because App Store’s Restriction.)…"
2 Thanks
I went out of town right when my LM4 arrived, so I haven’t even been able to play with it yet.
So to summarize this thread, and correct me if I’m wrong.
- The LM4 has somewhat useful CRI R1-14 measurement
- The LM3 is more accurate in everything else
Does that sum it up? Or was there another standout feature?
In my experience (meaning YMMV), depending on the lights.
- My low-medium CCT, high CRI lights read numbers that seem OK. Seem is the key word, because personally I don’t trust them considering what happens in number 2 below. Which makes LM4 a paper weight. And not even a good one since it’s so light
. - My high CCT, low CRI lights read CCT, Duv, Ra, R9 numbers that to me look like garbage (see my examples in previous posts in this thread). My LM3 has zero problem here.
Evidence with high CCT, low CRI lights were what I used to file dispute with Alix, which ruled 100% in my favor. Opple was forced to either give half refund, or full refund AND paying for return shipping.
3 Thanks
We have not yet seen any confirmation that R1-R14 are accurate. We do know that CRI R9 (Red), is Not accurate for Low CRI lights. We do not know if R9 is accurate for High CRI Lights…
LM4 calculates the DUV automatically, but I do not believe it is accurate.
The LM3 reads about -0.0023 to -0.0055 lower DUV than LM4. We have not yet seen any confirmation for which one is more correct.
Both meters seem to produce similar CRI Ra, CCT, and Flicker data.
the LM3 spectrum plot works, the LM4 spectrum plot does not work at all.
Here is an sw45k test:
in this example the DUV is -0.0042 (62%) lower on the LM3. I do not know which meter is closer to correct. Hoping someone w Xrite or Sekonic will share some LM3 and LM4 comparison results.
1 Thank
Your flicker free light is pretty flicker free.
- The graph shows ripple
- amplitude change is a bit over a 1/3
→ PWM is a square wave down to zero - Frequenz is 125Hz
Overall it’s not bad, yes the graph is not flat.
1 Thank
It shows as high risk according to that IEEE risk chart.
I measured a random assortment of lights on medium settings with stable sustained output. Typically 500-1500 lux depending on beam pattern. I’ll add some more measurements as I take them. I am somewhat limited in my ability test the effect of CRI because almost every low CRI light I have here is cool white.
| X-Rite i1Studio | Opple 4 |
|---|---|
| 3013K, Duv 0.0012, Ra 92.0, R9 58.7 | 2939K, Duv 0.0015, Ra 93.7, R9 65.9 |
| 4110K , Duv 0.0008, Ra 92.8, R9 64.3 | 4003K, Duv 0.0013, Ra 93.7, R9 67.1 |
| 3937K, Duv 0.0009, Ra 94.5, R9 69.7 | 3837K, Duv 0.0018, Ra 95.4, R9 73.8 |
| 5904K, Duv 0.0074, Ra 69.8, R9 -35.7 | 5430K, Duv 0.0037, Ra 69.3, R9 — |
| 6322K, Duv 0.0035, Ra 66.2, R9 -39.5 | 5898K, Duv -0.0012, Ra 65.2, R9 — |
| 6018K, Duv 0.0072, Ra 69.3, R9 -38.0 | 5523K, Duv 0.0032, Ra 68.3, R9 — |
| 4966K, Duv 0.0035, Ra 91.2, R9 69.4 | 4743K, Duv 0.0036, Ra 91.8, R9 78.6 |
| 5365K, Duv 0.0003, Ra 92.7, R9 68.0 | 5075K, Duv 0.0012, Ra 93.3, R9 65.9 |
| 4621K, Duv -0.0060, Ra 92.9, R9 72.6 | 4393K, Duv -0.0052, Ra 91.1, R9 69.3 |
| 4143K, Duv -0.0046, Ra 96.9, R9 84.3 | 4017K, Duv -0.0013, Ra 96.4, R9 75.6 |
| 4907K, Duv 0.0059, Ra 89.6, R9 45.5 | 4587K, Duv 0.0053, Ra 90.8, R9 65.9 |
| 5603K, Duv 0.0144, Ra 67.9, R9 -52.6 | 5234K, Duv 0.0106, Ra 68.9, R9 — |
| 2810K, Duv -0.0023, Ra 93.5, R9 72.3 | 2826K, Duv -0.0020, Ra 96.9, R9 78.1 |
12 Thanks
Thank you for sharing your measurements. It would be helpful if you could list the types of LEDs you tested.
Regarding the R9 measurements, I’ve noticed a wide range of numbers when measuring ceiling lights. However, I have found that all of my high CRI lights exhibit very high R9 values. Surprisingly, even the high CRI LH351D 5000K displayed R9 numbers above 95 in medium output, which I expected to be low. Additionally, I have noticed that many of my low CRI lights have also shown R9 values of 100.
Could you share more details about your setup? I am particularly interested in the distance between the light and the Opple.