Investigate the time synchronization among sensors

[moorepants]

See plots here: https://mechmotum.github.io/baby-vibration/#sync

This depends on #15 being completed.

An example that seems to show correct synchronization:

Many other trials have inconsistiencies and may not be syncronized. For example:

Some show how we probably have the sensor rotations incorrect, giving swapped signs for some of the sensors' longitudinal acceleration:

[moorepants]

@GabrieleDello can you look into this one also? There are a couple of sessions that may have time synchronization issues. I guess the simplest solution is to make sure the delta time is small and that we are selecting endpoints of the trial such that this would not matter.

[moorepants]

It seems that maybe only session 23 has this issue (although we need to zoom in on some sessions), and the difference is 0.5 seconds, so it does not matter for any analysis we are currently doing.