Okay, lets quickly start by saying I have been having issues with the F6X in OWS. After much experimentations and reading forums I have realized that its not just the hand that you wear your watch on, but in my case I need to wear the watch on the opposite side that I breathe on. I prefer to breathe to the left but usually alternate. Today I wore my watch on my right hand and only breath on my left and walla, my best track ever on the F6x Pro. Usually I get long straight lines (indicating no GPS for that duration) but today it looked good.
I also had the phone in my buoy and recorded the same swim on Strava.
So tracks look the same but the Strava (phone in buoy) distance is 4271 and on the F6X, 3333 yards, a big difference.
Okay, I then looked at those FIT decoders on the F6X swim file. What I saw (much better than past swims) is that GPS signal was there most of the time, however, it would periodically get lost and found again within seconds. Now, how does Garmin deal with that. Here is what happens:
While there is no GPS Signal: Garmin uses a fixed meters/second distance, (in my case it was .64m/s) which varies if you stroke count goes up/dwn, etc.
When GPS returns after being lost: This is where the majority of the error takes place. For the first 30 seconds or so after GPS has been re-obtained, the meters/sec value kind of ramps up slowly. So, for the first seconds its .01m/s, then .05 m/s, then .1 m/s for about 30 seconds, then it catches up to the right speed, in my case around .7 to .8 m/s.
So, while having no GPS it sticks to a fixed constant. But after re-obtaining GPS, for the first 30 seconds it thinks you are swimming very slowly which after a long swim adds up to much error. Here is a pace image that kind of describes what I am trying to describe. In tjhe picture below, every dip in pace is where I lost GPS. Thats when the watch takes 30 seconds or so to catch up to my real speed. The flat tops are when I had no GPS and it used a fixed value which is close enough to my pace but a little under. I believe that if Garmin can eliminate the 30 seconds of "very slow" pace after each GPS re-connection the total distance would contain far less error. Of course the real or better fix would be to improve on the GPS and make sure it is never lost, or rarely but that my not be likely. Okay, hoping Garmin reads this and makes some improvements.
