I don’t suggest you to use Opple Home 4.0.0 now because light master module still not being updated. I suggest you to use my windows app for now.
I’d talk with mobile app developers to do updates, but even they agree, it won’t be soon. they have their own schedule and process to do updates. app update doesn’t necessarily mean light master module getting updated.
3 Thanks
https://drive.google.com/drive/folders/1X2AgIluK49ep6BKINus8mJywQn21y4pw?usp=sharing
so difficult to upload files, google is inaccessible here and I used free VPN, speed is terrible.
5 Thanks
Do you want us to perform specific scenarios (test cases) and provide you feedback? Or it’s just to play with it a little?
I think it’s good enough now.
It’s appreciated if you guys could give me some feedback, let me know how to make it even better.
stevechang would you consider making your program open source and host it on github/gitlab
don’t think this will happen due to NDA (because of sensor manufacturer who makes great secret out of some information, something in this regard was described here) or something I think
2 Thanks
This thing is super cool, a fingertip-sized spectrometer with (for lighting) sufficient resolution. Sadly Expensive as hell (450$ as far as I could find).
once proposed to open source it, got rejected, signed NDA with relevant parties. plus I am just an engineer hence I don’t have the capacity to decide on this “company level” thing.
4 Thanks
one hamamatsu employee at hamamatsu China sent me an evaluation board to test it last year actually, the chip is not suitable, 1. the one he sent me behaved weirdly, he said in this specific field ams perhaps is a better option, hamamatsu is good at other specific fields 2. the form factor is not suitable for us.
1 Thank
somebody ever tested flashlights using the Windows app? is it accurate enough for your testing?
I can test some LEDs with windows software and my spectrometer. Which LEDs (CCT / CRI) should I test?
2 Thanks
I think as many as possible:-)
several days ago I posted v1.1 in google drive to replace the v1.0, you should use v1.1, I think my testing previously is not as trustworthy as you from other people’s perspective.
and you can actually post your previous testing results too, you said it’s awful. btw, although the sensor only has 8 VIS channels, and I think we’ve already achieved better results, we are still try to make it even better.
thank you very much.
4 Thanks
I only tested the Opple smart home app (one of these modified version), I can compare these with the v1.1 windows software.
I think I will these LEDs of different shape, CCT and CRI:
- Luminus SFT-40 6500 K (“slightly green”)
- Yinding 5050 CRI 95 6500 K
- Nichia 519A 2700 K R9080
- Cree XHP70.3 HI 70CRI 4000 K (“rosy”)
- Fireflylite FFL505A 3500 K 95CRI (“very rosy”)
4 Thanks
I took a bunch of measurements with the new Windows app today - works like a charm! Scaling for different window sizes is a bit broken, but the measurements are fast, all values are there and the reconstructed spectrum graph is super nice to see.
Thank you Steve!
I will post some of them as part of my Wurkkos HD10 review (4000K 90 CRI LEDs, apparently), so once other reviews of the same light with professional equipment are published we can compare results.
So far all measurements I took matched my expectations, unlike the Android app which sometimes produced weird results.
Thank you for your motivation and persistence!!
1 Thank
actually I’d like to thank you guys for your patience, will continue to improve the software, this is something that is within my control.
7 Thanks
Your willingness to continue working on improving the software is why I have not given up on the LM4, or the entire idea of being able to measure my lights with equipment that cost less than my car. I can’t speak for anyone else, but I definitely appreciate you going above and beyond to resolve the shortcomings of the LM4 and trying to make it the best it can be. And your willingness to engage with the community and listen to feedback is admirable. I hope that your superiors share the same views and reward you in ways more substantial than our patience and gratitude. You deserve it.
6 Thanks
Honestly I had given up getting a good software for this and decided to treat it as a “might be partially useful” toy.
Then you came along, and I started having hope of having a useful tool after all, and now we reached the point where I can actually take measurements with the LM4. I am very happy I bought the LM4 and waited. It’s nice to use now.
You went above and beyond what the company tasked you with, and made a lot of people very happy out there.
4 Thanks
Here they are:
Looks pretty good, although there are some smaller differences in reconstructed spectra and therefore the results. Huge improvement since the Android app versions, so good job @stevechang ! 
Thanks a lot for not let the LM4 abandon, but to constantly improving it despite the limited resources and help from others at Opple!
Would be nice to see some exporting and picture saving features in this software for easier sharing of results. also a zoomable CIE1931-chart and a dedicated picture saving function only for CIE1931 would be great. (And a version number under “about” could be handy for checking if the newest software version is installed.)
7 Thanks
My personal wish would be exporting the report to CSV, where it exports both the spectrum (I guess it is saved as a formula?) and all the result values, so I can store them in small text files and create new diagrams in a data evaluation software of my choice at a single click
1 Thank
Pretty sure the formula for reconstructing spectrum is under NDA, the top secret Chinese thing 
But CSV export in general would be great.