On reflection, I think you’re right… 2.3 seconds is not an approximation of 1 second. I fell in to the trap of thinking from the firmware perspective and letting the user perspective slide too much. I think of the sampling interval as the sleep time between samples, disregarding the sampling time itself, but the majority of users will have your expectation. Consider “making time between samples in log samples correspond to selected sampling time” on the to-do list.
Regarding when to start a new logging session – right now the firmware should generate a new log file every time you switch logging off and on again. Is this what you mean?
Thanks again, you’ve given a lot of feedback over the course of the project and it’s been really helpful!