Also gave me a new (Strava) 5k estimated best of 24:48.
Strava also seems to think I stopped for 1m40 during the run and gave me a moving time of 28:32, but that's a load of old shite as I didn't need to slow once.
The 24:48 best 5k looks legit, I wrote my own program to parse the .tcx file. Blimey.
As for the lower moving time. Ah, the tcx file is missing lat/lon points for random trackpoints (no idea why, they're randomly throughout the file and nowhere near the bits where the GPS should lose signal). See the middle trackpoint of the 3 in the attached image. It's 5k into a run in an area where there should be no loss of signal.
Removing the Trackpoints without positional data and reuploading has fixed it, it correctly has a 30:12 moving time. I'll try adding in positional data by interpolation so that I keep the HR data that I'd be losing by dumping those dodgy points.
The 24:48 best 5k looks legit, I wrote my own program to parse the .tcx file. Blimey.
As for the lower moving time. Ah, the tcx file is missing lat/lon points for random trackpoints (no idea why, they're randomly throughout the file and nowhere near the bits where the GPS should lose signal). See the middle trackpoint of the 3 in the attached image. It's 5k into a run in an area where there should be no loss of signal.
Removing the Trackpoints without positional data and reuploading has fixed it, it correctly has a 30:12 moving time. I'll try adding in positional data by interpolation so that I keep the HR data that I'd be losing by dumping those dodgy points.
1 Attachment