Running in a tunnel caused issues in the total distance computation

tuxella

Hello,

I ran a couple of times in a tunnel with my watch (S9 Baro) and it caused the total distance to be very off (so wrong that strava flagged it).

Here are the activities, on strava and QS :
30/03/2021:

09/03/2021:

When looking at the GPX file and re-computing actual distance between points, the distance is ok (7.74 km, when the friend I ran with measured 7.72 with his watch).

My current working theory is that the step length remembered by the watch is too long.

How can I re-calibrate this length? (The track of the 30/03 was done after the firmware update, so after a soft reset AFAIK)

Dimitrios Kanellopoulos

@tuxella the recalibration happens a the time with gps based workouts.

If you run in a tunnel that is straight then the GNSS distance from gpx recalculation (or check it on qs) will be ok and right. If the watch reports more it indeed does not use so we’ll the fused speed or you look many times at your watch.

tuxella

The tunnel more or less follows the yellow line bellow, which I think is pretty straight.

/var/folders/8t/5y2nkcwj2bs623k4l75359xm0000gn/T/com.sindresorhus.Shareful.Copy/87060D8D-1FCC-4BA8-85E8-408B62B00074/Untitled.png

See also the openstreetmap routing:

https://www.openstreetmap.org/directions?engine=graphhopper_foot&route=48.85891%2C2.33983%3B48.86052%2C2.33101#map=17/48.86011/2.33463

I don’t think I looked at my watch more than once or twice under the tunnel (which is ~700m long). The cadence and speed reported on points while I was under the tunnel look ok to me in QS.

Can I do anything to help the watch (using fusedtrack or not) for my next run?

sky-runner

@tuxella I suspect the watch may be double counting the tunnel distance in this case - the first time from accelerometer and the second time from GPS when it reacquures the GPS signal. The easiest way is to download the FIT file and inspect the distance over time in it, and see if there are any sharp jumps in the distance. In my experience Ambit3 had exactly the same issue, which I observed several times.

Dimitrios Kanellopoulos

@sky-runner we have fixed this issue quite some fw before.

tuxella

This “double counting” theory could explain the results I get.

I analysed the data in the Suunto fit file from the 30/03/2021 :

The first time in the tunnel (difference between the first and the last point without coordinates):

Distance: 0.87 km
Time: 4:15
Speed: 12.28 km/h

On the way back (in the same tunnel):

Distance: 0.88 km
Time: 4:19
Speed 12.23 km/h

If I add these distances to the distance calculated from GPS coordinates of each point:

7.74 + 0.87 + 0.88 = 9.49km

Where the total distance in the fit file (and reported in Suunto App, QS, Runalyze, Strava…) is 9.58 km

Dimitrios Kanellopoulos

@tuxella can you make the links of qs public ? You can easily see if it’s that bug which was fixed

Currently I cannot open those links

tuxella

@dimitrios-kanellopoulos They should be public now

Dimitrios Kanellopoulos

@tuxella are you using a pod ?

tuxella

@dimitrios-kanellopoulos No, this activity was done with just a Suunto S9 and the HR belt

Dimitrios Kanellopoulos

@tuxella I’ll take a better look when I am on my pc.

I do see the issue.

tuxella

@dimitrios-kanellopoulos Cool

Have a nice week-end

ddausw

@tuxella
I have the same problem with the s9b after the last FW update. The tunnel length is counted twice. I tested with Glonass as well as Galileo.

Dieter

MiniForklift

Nice pace! Just gave you kudos

tuxella

@miniforklift Thanks

tuxella

@dimitrios-kanellopoulos Hello

did you have the opportunity to check on this?

Dimitrios Kanellopoulos

@tuxella I reported a bug or the watch guys.I think this new FW re introduced the issue

Dimitrios Kanellopoulos

@tuxella I would like some suunto app links

tuxella

@dimitrios-kanellopoulos

For the 30/03/2021: https://www.suunto.com/fr-fr/move/thomas430851/606310cf949ba71fa290dc8a

For the 09/03/2021: https://www.suunto.com/fr-fr/move/thomas430851/60476ac697a7b66e0c4c5d8a

tuxella

@dimitrios-kanellopoulos I had this issue twice: on the 30/03 and on the 09/03.

On the 09/03 I wasn’t using the “new firmware” (which was released on the 30/03 AFAIK)