This object is in archive! 

Data imported from Garmin is incorrect

Ádám Berhidi shared this problem 4 years ago
Not a Problem

Hi,


So far I have been recording hikes with my phone but recently switched to a Garmin watch which measures elevation too.


My issue is that when I am importing the .gpx file from the watch most date are off in Locus Map, compared to what I see in Garmin Connect. Distance, altitude, elevation gain are considerably off, but also heartrate as well (although only slightly).


I am wondering that Locus is recalculating everything based on map data? How can I avoid that?


Thanks,


Adam

Replies (4)

photo
2

Hello Ádám,

may you please share some exported track + screenshot of values you see in Garmin web page? Usually, best is to compare data against some third (or fourth) application, to get a better idea, whether the issue is on Locus Map or Garmin side (which is exactly what I will do with your track as the first step).

Menion

photo
1

Hi Menion,


GPX is attached, I assume Locus may adjust based on GPS data and topography? Also realized my watch had 3D distance feature switched off which may also impact calculations as I was hiking.

Below is a comparision of the data I have, note that the Connect values are non-adjusted. If I flip to GPS adjusted in Connect, it changes the elevation data considerably, but makes no change to the overall distance.


Garmin Connect / Locus imported


Distance: 15.58km / 17.76km

Elevation gain: 673m / 692m

Tracking time: 5:24:01 in both

Moving time: 4:00:45 / 4:48:33


Thank you,


Adam

photo
1

Probably connect reports automatic filtered values ? (See Mapsource)


The track has a few locations where was a pause, strolling around, that are contained into the total unfiltered distance.

Filter: A hard task for automatic filtering reports = The result is a lotery. (Such a track should be pre-cleaned manually)


(imo)

Many track records have a fairly low quality, and are just sufficient to have a rough indication.

Due to too low quality, so not worth breaking his mind about, because of some statistics.


UNFILTERED.


Gpx editor: 6255 points 17,2 km.

Mapsource: 6255 points 17,2 km.

Basecamp: 6255 points 17,2 km.

Gps Track Editor: 6255 points 17,15 km


FILTERED.


Mapsource:

auto filter (set 4 max) 1912 points 16,2 km

auto filter (set 3 med) 1346 points 15,8 km

auto filter (set 2 med) 791 points 15,1 km

auto filter (set 1 min) 616 points 14,7 km


Gps Track Editor:

Filter (local inconsistencies): 5253 points 15,69 km.


In attachment find a clean trackrecord, where the recorded distance is compared with the design distance.

However most (uncleaned) trackrecords are just not reliable enough to withstand such a compare test.

photo
1

Thanks Willy for very precise test!

Can't say more. Locus Map really compute distance from coordinates without any extra filtering.

For recording over Locus Map, it is possible to enable filtering that may "improve" recorded values, but for imported tracks/routes, no filtering is possible.

Surprising is difference in elevation, because this value usually makes biggest difference.

So generally, values in the app are not problem (as 0709 confirmed by precise tests), but a result of interpretation.

Jiří M. aka Menion

photo
1

Thank you both.


I am fairly new to the world of GPS watches, but I am long time user of Locus as is, based on data from my mobile phone. So apologies for the dumb questions :)


Which is the correct distance then?


A have the Locus app which I really like and a Garmin Instinct, how can I leverage the best results then? Can I somehow record directly from the watch to the app? I tried to pair in the Locus settings but no joy.


Thanks,


Adam

photo
1

What is the correct distance then ?

In my opinion, no GPS distance measurement or interpretation afterwards by filters is correct.

One hardware (free setup, view of satellites) may provide slightly more correct data than others.

You have to discover for yourself what you prefer. I don't break my head about it anymore.


Correct distances by bicycle? Use a CALIBRATED wheel revolutions counter as the most reliable indication.

Correct distances on foot? That is more difficult ... walking around with a surveyor's 'measuring wheel' does not seem appropriate to me, but would give a correct result.


10 devices> 10 different results.

It has to be said that a GPS watch on the wrist does not improve accuracy due to the rather unfavorable location for GPS.


Test and results with different devices?

See one here: https://www.navigation-professionell.de/tracks-garmin-etrex-30x-test/


This confirms, a GPS is a fun toy, but not a precisely traveled distance measuring device.

photo
1

Thanks... In this case the difference is too much for this short trip, it is more than 1.5km.

Maybe too bad reception? I remember my phone did not pick up any GPS for many kilometres as we were in a valley.

Not sure about the filters/interpretations used by the watch for the last one but will check with Garmin.

I checked my previous 2 hikes, in those cases the difference was much less significant, just a few hundred meters.

Maybe the last hike had too much stops (we were taking many photos) so Garmin Connect may have truncated it.

I did a running yesterday, which is a 100% match in Locus and Garmin connect.

Thanks,

Adam

photo
1

Just some feedback on the original post...looks like Garmin was automatically neglecting all the short stops for photo making, etc, because i was too slow, not recording the distance there as a result.. There was a setting in the watch for data recording, I changed it from "intelligent" to "second by second".


And now Garmin actually gives more distance (19.14 km) while if imported to Locus, it is only only 18.62 for a recent trip.

Out of curiosity, I also recorded with on Locus using my phone, getting 17.10 km.

https://connect.garmin.com/modern/activity/4983583584


Weird :D


Cheers,


Adam

Replies have been locked on this page!