If you have a windows machine, check above for a link to the software @stevechang has been developing. By the reports it makes results much closer to reality.
It seems that the app has more impact on the strange results than the device itself.
Welcome
portability is an issue for windows software,
currently I spend time on other stuff, secondly I have to port C/C++ code to JavaScript cuz Android/iOS run JavaScript code, thirdly I have to make the PM consent to my update request which also comes not easy.
I have full confidence that the new algorithm outperforms LM3, LM3’s results is just “happens to be true”. and the Ra of daylight is not necessarily always 100.
Thanks @stevechang for all the great work you did for the LM4. You are the reason why I bought one. Take care.
Hello,
so, I am looking for such a device. Is it now at a level with good accuracy and reliable results?
Are there any alternatives in a price range that is not too expensive?
Thanks!
It seems to work mostly fine now when using the windows app that Steve has made, with good accuracy for the price.
The next step up would be a used Xrite Colormunki Photo, but those are a $100-150 price range.
There is also this for ~60-80 USD on Aliexpress, a passion project from a Chinese hobbyist. I ordered one and it is on the way here. Curious how well it works. It is a proper diffraction grating + monochrome CMOS sensor spectrometer, but built with very low budget materials to allow the price tag. Will see how it works once it is here.
thank you for your information, I am in contact with the hobbyist.
he said he chose monochrome CMOS sensor instead of filter-based sensor like as7341 is that as7341 has a extremely low resolution. I may suggest this to superior.
Based on the description the Bayer array is removed from the CMOS sensor. There are a few ways to do this, people make a living out of it. Most crude would be mechanical / scraped, but it can be done chemically as well (a slow, manual, painstaking multi step process involving solvents and cleaners to remove residue).
The question is, how consistent this process is from sensor to sensor, and whether inconsistencies can be known / calibrated for.