Feature request: Precise “start recording at”

[lobradov]

Current “start recording at” option allows definition of hour and minute to start recording (with assumption of 00s and 00ms). With multiple cameras that have synced time (either via codes or via GPS), I noticed few seconds of difference in when recording actually starts. I also noticed that start is not consistent on the same camera - recording would sometime start with few ms of delay (so almost right on time), sometimes would take 2-3 sec.

Hoping for improvement to start recording at exactly defined hh:mm(:00:00).

[dnewman-gpsw]

It is a good idea, and the camera tries to work that way most of the time. A command !S <- start at a 24 hour clock time, will boot the camera before the required time, and wait for the appropriate second. But it doesn't attempt for anything more precise than that (RTC doesn't have subsecond precision.) However, sometimes the camera wakes a few seconds late, in these cases the capture starts immediately, but making it more out of sync with any other camera attempting the same scheduled start. The feature is typically good for 1-2 seconds accuracy, I want more also, just a few more technical hurdles on this one.

[dnewman-gpsw]

Idea for this, if you want all cameras to start at noon, !12:00S would normally have been the command. However you can do !11:59N!12:00S instead. This should improve the precision a tad. When camera boot late, it is normal only 30-40 seconds late. The precision improves as the time jump shortens. The idea is still good to investigate further. Leaving open.