The LM3 sensor uses the AS7262, which has 6 channels in the visible wavelength range, while the LM4 sensor features the AS7341 with 8 channels. Although the LM4 supposedly has better hardware, the original LM3 development team left Opple without a smooth transition (see here). LM4 was then developed by a team lacking expertise in color science. Since its launch, we’ve noticed that LM4’s measurements have been inconsistent, suggesting the product may have hit the market before the app was fully refined. Steve mentioned that no one in the development team specialized in colorimetry, so he had to self-train to enhance the app.
Here are the specs of the sensors used in LM3 and LM4
LM3 - Osram AS7262 (6-channel visible sensor)
See here for LM3 sensor id
LM4 - Osram AS7341 (8-channel visible sensor)
See here for LM4 sensor id
The sensors in these devices are nowhere near the quality of those used in even a moderately acceptable spectrometer, which makes their accuracy inherently limited. To illustrate, the sensors in LM devices can be comparable to a SD 360p camera, while those in a proper spectrometer are comparable to a 4K camera. While a 360p camera is certainly better than having no camera at all, it lacks the resolution needed for detailed or precise results. I personally think we do not need such a high resolution or accuracty for our purpose (testing flashlights) and that is why I thought LM devices were useful.
I’m disappointed to see that Opple has decided to discontinue its LM product line. These devices provided a useful and cost-effective way to assess the quality of lights.
Thanks to the users here, we can at least make the most of what we already have.