Opple Light Master 4 discussion thread (new 2023 model)

ebastler · May 19, 2024, 7:49pm

Oh, not the reconstructing formula - just the formula or points for whatever we see rendered. So not the formula that takes the sensor values and constructs a spectrum, but the end result which the software uses to render the pic on the screen. Basically, I would love being able to re-draw those graphs in an external software from whatever is given in the CSV - be it a formula that describes this curve, a lot of individual points, or some important points and I have to overlay a smoothed curve.

Here’s the spectrum of a Wurkkos HD10 btw, CSP2323 in 4000K. 95 CRI according to the LM4.

koef3 · May 19, 2024, 10:29pm

Ok you just mean the values for the wavelength itself.

ebastler · May 21, 2024, 10:10pm

I took some measurements - B35AM 3500K R9080, XHP70.3 4000K R70, SFT40 5000K R70. All looks good, mostly as expected. The UI of the tool is a bit buggy with overlapping elements on my screen, but gets the job done

The DUV graph seems broken, however, since all my 3 LEDs are around 10000K in it. Spectrum and calculated values look good.

stevechang · May 22, 2024, 1:13am

sorry, only tested on 2 laptops, so didn’t aware of different screen resolution ratio, need to plan UI layout.
btw, I am actually now working on the 5th version, and it’s not for release. so I’d spare some time occasionally to optimize the windows software. for the 5th, the bluetooth chip is to be replaced, and it would come with a screen and support distance measuring. I am working on the embedded software as well, not familiar with the chip at all, so learning and working on it now.

jerv · May 22, 2024, 7:13am

Your 3440x1440 window looks a bit different from my 2560x1440, and other shots here seem to be on 1920x1080 that have a lot less whitespace than either of us do with our 1440 monitors. The graphs don’t overlap on a 16:9 but those metrics require a window that’s as wide as the screen to show. And for you and I, that’s a lot wider.

Here’s my 5000K B35AM DM11 to show you the UI differences and another emitter

Easy to miss. There are not many laptops with a screen 2560 pixels wide, let alone 3440. A lot of laptops are 1920x1080 or smaller. And a lot of people use 1920x1080 on their desktop computers as well.

stevechang · May 22, 2024, 7:42am

maybe should use a fixed size window, for example 800*600.
how about the accuracy of your measurement, because there is no comparison with professional spectrometer.

koef3 · May 22, 2024, 8:18am

I think dynamic windows sizing is better. It is not good to use fixed values in programming since these could make problems in some (more special) use cases.

stevechang · May 23, 2024, 5:23am

on my personal laptop the UI is like this, the screen resolution is different from my office PC.

AEDe · May 23, 2024, 11:29am

On my laptop I can not see R4 R8 R12… But I check several led with big difference LM4 Android aplication - ColorMunki. and other 11 Rs are very close to ColorMunki measurements now. Can not find difference more than 5 units.

stevechang · May 23, 2024, 12:49pm

Could you go into more details? I can’t understand you fully.

jerv · May 23, 2024, 3:20pm

They don’t show up on 2560x1440 either (see above). It seems like the UI scaling works right only with 1920x1080

ebastler · May 23, 2024, 3:22pm

My notebook internal screen is 2880x1800px (16:10), it works nicely there too.

AEDe · May 23, 2024, 3:59pm

It seems depend on screen ratio my laptop has 3:2.

stevechang · May 24, 2024, 1:33am

let’s first verify the accuracy, then I’d optimize the software incrementally, I think koef’s suggestion is good, for example feature like supporting export measurement results.

AEDe · May 24, 2024, 4:16am

Ok . If needed I could post comparison ColorMunki and ML4, but without R4,R8,R12.

stevechang · May 24, 2024, 8:51am

appreciated

AEDe · May 24, 2024, 6:43pm

1)E17A 1850K
Color Rendering Index (Ra) = 95.0 [ R9 = 94.1 ]
R1 = 95.4 R2 = 96.4 R3 = 97.5 R4 = 93.0 R5 = 93.5 R6 = 90.4 R7 = 96.7
R8 = 97.2 R9 = 94.1 R10 = 94.5 R11 = 86.5 R12 = 83.2 R13 = 94.7 R14 = 96.7
Television Lighting Consistency Index 2012 (Qa) = 94.9
IES TM-30-15 Rf = 93.84 Rg = 100.10 CCT = 1867 Duv = -0.000305

2) E21A 2000K
Color Rendering Index (Ra) = 94.9 [ R9 = 98.1 ]
R1 = 94.1 R2 = 95.9 R3 = 97.4 R4 = 91.9 R5 = 92.5 R6 = 90.3 R7 = 98.0
R8 = 99.2 R9 = 98.1 R10 = 93.3 R11 = 85.8 R12 = 83.5 R13 = 93.6 R14 = 96.8
Television Lighting Consistency Index 2012 (Qa) = 95.8
IES TM-30-15 Rf = 94.12 Rg = 102.20 CCT = 1975 Duv = -0.000802

3)Nichia 519A 3000K
Color Rendering Index (Ra) = 95.5 [ R9 = 91.4 ]
R1 = 98.9 R2 = 97.4 R3 = 92.1 R4 = 93.7 R5 = 98.1 R6 = 95.7 R7 = 94.3
R8 = 93.7 R9 = 91.4 R10 = 93.3 R11 = 93.7 R12 = 89.9 R13 = 99.5 R14 = 95.1
Television Lighting Consistency Index 2012 (Qa) = 97.8
IES TM-30-15 Rf = 93.94 Rg = 104.31 CCT = 3287 Duv = 0.000106

4)Samsung 351D 3500K
Color Rendering Index (Ra) = 94.0 [ R9 = 77.9 ]
R1 = 94.0 R2 = 94.1 R3 = 93.3 R4 = 95.7 R5 = 93.0 R6 = 91.8 R7 = 97.9
R8 = 92.4 R9 = 77.9 R10 = 85.8 R11 = 95.5 R12 = 77.7 R13 = 93.5 R14 = 96.0
Television Lighting Consistency Index 2012 (Qa) = 97.6
IES TM-30-15 Rf = 93.22 Rg = 99.43 CCT = 3364 Duv = 0.004517

5)Xhp50.2 90cri (zl h600c)
Color Rendering Index (Ra) = 93.0 [ R9 = 71.3 ]
R1 = 92.9 R2 = 94.6 R3 = 95.0 R4 = 93.2 R5 = 91.9 R6 = 91.4 R7 = 96.2
R8 = 89.2 R9 = 71.3 R10 = 86.4 R11 = 92.5 R12 = 73.2 R13 = 93.0 R14 = 97.2
Television Lighting Consistency Index 2012 (Qa) = 94.7
IES TM-30-15 Rf = 90.71 Rg = 98.87 CCT = 3935 Duv = 0.002107

xpl hi 3000K

image2242×1147 185 KB

Color Rendering Index (Ra) = 83.8 [ R9 = 21.3 ]
R1 = 82.6 R2 = 89.6 R3 = 95.4 R4 = 82.4 R5 = 82.1 R6 = 86.2 R7 = 85.9
R8 = 65.9 R9 = 21.3 R10 = 75.8 R11 = 80.6 R12 = 68.4 R13 = 83.7 R14 = 97.5
Television Lighting Consistency Index 2012 (Qa) = 71.9
IES TM-30-15 Rf = 82.90 Rg = 98.12 CCT = 2988 Duv = -0.000200

7)Xml2 4500K (zl 602w)

Color Rendering Index (Ra) = 76.2 [ R9 = 7.3 ] (Caution)
R1 = 77.4 R2 = 79.9 R3 = 78.7 R4 = 77.1 R5 = 76.6 R6 = 70.4 R7 = 81.7
R8 = 67.8 R9 = 7.3 R10 = 50.1 R11 = 74.3 R12 = 50.5 R13 = 76.5 R14 = 88.0
Television Lighting Consistency Index 2012 (Qa) = 48.1 (Caution)
IES TM-30-15 Rf = 69.44 Rg = 98.95 CCT = 4809 Duv = -0.010448

ZL A6 luxeon V 5000K

image2263×1135 188 KB

Color Rendering Index (Ra) = 71.7 [ R9 = -26.7 ]
R1 = 69.2 R2 = 75.3 R3 = 81.1 R4 = 72.8 R5 = 70.3 R6 = 67.3 R7 = 79.9
R8 = 57.5 R9 = -26.7 R10 = 43.5 R11 = 71.1 R12 = 47.1 R13 = 69.0 R14 = 89.8
Television Lighting Consistency Index 2012 (Qa) = 45.4
IES TM-30-15 Rf = 70.14 Rg = 95.70 CCT = 4785 Duv = 0.002139
XRE ZL h50

image2239×1121 178 KB

Color Rendering Index (Ra) = 77.0 [ R9 = 1.6 ]
R1 = 76.7 R2 = 80.2 R3 = 79.0 R4 = 78.2 R5 = 76.7 R6 = 70.6 R7 = 84.6
R8 = 70.1 R9 = 1.6 R10 = 49.7 R11 = 74.4 R12 = 48.3 R13 = 76.6 R14 = 89.1
Television Lighting Consistency Index 2012 (Qa) = 54.5
IES TM-30-15 Rf = 72.73 Rg = 94.76 CCT = 7036 Duv = -0.005160
KSF phosphor lamp 3000K

Color Rendering Index (Ra) = 91.1 [ R9 = 79.1 ]
R1 = 95.3 R2 = 91.3 R3 = 83.0 R4 = 92.3 R5 = 92.3 R6 = 88.0 R7 = 93.1
R8 = 93.2 R9 = 79.1 R10 = 75.0 R11 = 92.5 R12 = 68.3 R13 = 93.4 R14 = 89.8
Television Lighting Consistency Index 2012 (Qa) = 80.0
IES TM-30-15 Rf = 86.88 Rg = 104.01 CCT = 3096 Duv = 0.002148

stevechang · May 25, 2024, 12:40am

thank you very much. so the conclusion is?

AEDe · May 25, 2024, 5:04am

ML4’s CRI is very close to ColorMunki. Difference no more than 2.
Ri are also close to each other .
Errors in R9 are bigger then in other Ri but still acceptable.
Works good even with KSF led
Conclusion : this is definitely the best result I ve seen for LM4 , but R4,R8 R12 needed to be sure.

stevechang · May 25, 2024, 12:56pm

I will adjust layout, r4/r8/r12 actually does exist on the window, just outside of your screen.