Richard’s track.
This watch uses the MediaTek MT3333.
Firstly, stating this was entirely a consequence of running the data through GPSBabel, and serves as a warning that tools can do unexpected things to your data!
The end result was an apparent 25.36 knot spike @ 15:52:44 followed by a big 44 knot spike, following a crash. These spikes were NOT in the original FIT data.
The standard “max acceleration” filter excluded these from the GPSBabel results since “Sats”, “HDOP” or “SDOP” are not present.
Important note: This spike is not present in the Doppler speeds of the original FIT. It was only introduced by GPSBabel and is derived from positional data.
You can find all of the tracks on GitHub under sessions/contacts/bear/tracks.