I have been trying to assess exactly that. Sadly, it would appear that those numbers are equally difficult to relate to actual remaining capacity. Don’t forget, Blink has that number too, and they can’t do it, well. All the fancy algorithms in the world don’t help when, as Joel explained, the batteries effectively go, good, good, good, good, good, dead.
This however is not a failing of the batteries, and is in fact a desired trait. Having batteries that have what is described as a flat discharge curve is highly desirable property, and lithium are a prime example. Nothing to do with them being cheap, they are brilliant batteries. They are cheap because of the huge volume, that’s all. The problem here is that this almost constant voltage makes it very difficult to determine their remaining capacity based on voltage alone.
It’s similar with Lithium ion, but in your expensive phone, with battery power to spare, special hardware, and processing power to spare, it can use multiple methods, voltage, coulomb counting, internal resistance, discharge/charge profile tracking, and some other clever tricks, to come up with a more accurate remaining capacity, which is why it is so much more reliable. Even here, you sometimes get odd behaviour from the battery indicator. The key though is that it’s not the quality of the battery, it is the quality of the device assessing it, and a smartphone has a much bigger armoury.
I don’t think it’s feasible for the Blink camera to be equipped with a similar arsenal. It would increase complexity, cost and power consumption, the last of which is a target goal.
Going back to the most recent question though, when I performed the last battery change, I ‘was’ notified of low battery. First time ever. Mostly I have to change due to obvious erratic behaviour, but the indicator in the app still shows good. This time the camera continued to work, for a good few clips, of around a minute total, overnight. Didn’t want to change them in the dark and damp, so did it next day. The last clip however was cut short, when there was clearly still motion, and it should have recorded it all.
As soon as I got the low reading in the app, I checked in larume84’s app and this particular set showed 120, as recorded on the Blink server.
Most others show between 150 and 160, dependent on age. Some are down to 145, and still working fine.
120 was way out, and clearly too low, but catching it as it goes from 150 ish, to 120, will be a tricky task, because it will be quick, for all the reasons described above, so I’m not sure larume84’s welcome inclusion of this value really helps. I’m going to continue to monitor them, and see if I can get anything useful from those values, but the lack of other data makes it just as tricky for a person, as it does for the Blink camera or server.