Welcome To Fetcheveryone

Our awesome training log doesn't hide its best features behind a paywall. Search thousands of events, get advice, play games, measure routes, and more! Join our friendly community of runners, cyclists, and swimmers.
Click here to get started
Already a Fetchie? Sign in here

Garmin

188 watchers
18 Oct
11:49am, 18 Oct 2021
15781 posts
  •  
  • 0
larkim
MMM that makes sense other than the fact that every other run I've done in recent memory has not had the "short" laps. It makes me nervous when my watch measures a mile and strava measures that same distance and they get a different distance!

Maybe it'll be a one-off.
18 Oct
12:10pm, 18 Oct 2021
20069 posts
  •  
  • 0
flanker
strava works on the raw GPS data (or at least always used to - not sure with all the recent changes) and calculates the distance itself rather than using reported value (as does Fetch iirc) and it will depend on the algorithm they use, if any smoothing is applied, the precision it works to, etc.

I see it quite often, but as I don't use strava (or Fetch come to that) as a log have never got around to looking into it.
18 Oct
12:23pm, 18 Oct 2021
15783 posts
  •  
  • 0
larkim
Yes, but the laps info in a .fit file shouldn't get truncated like that.

Viewing the .fit file, the laps all show as being 1609.34m long (correctly) with precise durations and whilst I can accept that for presentation on the "overview" page of strava it has laps which are simply generated from the raw data, the "laps" page should simply report the .fit file laps info, rather than use the start / end points for the laps and then do a recalc.

Bizarrely, even the times to complete each lap are different to the .fit file table, and it's not even a consistent rounding up / rounding down pattern.

.fit file is 7:13.3, 6:55.9, 6:55.6, 6:49.4, 6:49.1
strava laps are 7:14, 6:54, 6:54, 6:49, 6:47

And surely those times should just be taken from the .fit file laps dataset?

It doesn't matter really, of course, but I do like to get excitable over these sorts of things.
18 Oct
12:28pm, 18 Oct 2021
15784 posts
  •  
  • 0
larkim
Reddit has come to my aid.

Apparently if you crop a run it can skew the resulting laps, and I'd cropped out a couple of minutes of standing around time at the end after I'd mistakenly hit resume.

Just re-uploaded the original .fit file and the laps match properly.

Mystery solved, and knowledge added to the databank!
18 Oct
4:10pm, 18 Oct 2021
20071 posts
  •  
  • 0
flanker
Ah, that sor of makes sense, although I'd question the underlying logic being applied. Worth knowing*

* probably not really, but it interests me anyways :)
18 Oct
4:15pm, 18 Oct 2021
15786 posts
  •  
  • 0
larkim
It is curious that they over-write a fixed piece of data, definitely. It's as if they are compressing the laps into the remaining space for some reason.
18 Oct
4:19pm, 18 Oct 2021
15787 posts
  •  
  • 0
larkim
Curiously, if I re-upload and crop the activity waaaay back, it does seem to properly remove laps that couldn't exist (e.g. laps 5 and 6 if I trim it back to being a 4.x mile run), but still adjusts the distances of those laps down a bit as per the original query.

Goes to show why we should continue to use Fetch and Fetch alone :-) when these multi-million dollar companies can't get simple things like that right!!
18 Oct
4:54pm, 18 Oct 2021
14264 posts
  •  
  • 0
Badger
Can you also crop off the start? It might just automatically be evaluating whatever is left, oblivious to whether the first part should be right or not.
18 Oct
4:58pm, 18 Oct 2021
15788 posts
  •  
  • 0
larkim
Even more odd if I do it that way round. It leaves part of whatever lap the crop starts in (in the case I just played with, 0.04 of a mile) and then all of the others get the same peculiar 0.99 treatment.

Bet bet Lord Fetch wouldn't have designed it like that...
18 Oct
8:12pm, 18 Oct 2021
10401 posts
  •  
  • 0
Sigh

About This Thread

Maintained by fetcheveryone
For all questions and comments about Garmin devices, including importing to Fetch.

Related Threads

  • garmin
  • gps
  • tech

Back To Top