This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Why my Forerunner 245 distance measurement are really bad?

Hello

I own a Forerunner 245.

It has firmware updated to latest.

I noticed that the difference in measurement between my device and the one of a friend of mine (smartphone) is really different.

Yesterday he measured 11.7Km while I measured 7.9Km, we were walking together.

My watch has GPS smart measurement turned off so it read a point every second.

The walk was along the coast with 250m down and up (descent/ascent).

I work with cartography (Geographic Information Systems) so I know how to measure real 3D distance of a path.

I exported the GPX file of the activity, imported as points in my GIS software (6516 points).

I then reproiected them as EPSG 32633 (UTM33N-WGS84) to have cartesian measurement, not angular (meters not degrees).

GPX is Lat/Long WGS1984 datum, UTM33N-WGS84 is X/Y WGS1984 datum so the passage between the two coordinate system don't introduce much error.

Then I reconstructed the path that is UTM33N-WGS84 (meters) and I calculated the 3DLength of it. It is 10.1Km (not 7.99 as reported by Forerunner/connect...)
I then tried to disable elevation correction in Connect and I remade all process (export GPX (is somehow different), import, reproject, create path, calculate 3d distance). The distance is 10.05Km, similar to the other one

I then tried to use a "high resolution" digital terrain model (10m x/y resolution) official from our government. I corrected the elevation of my GPX file with those from the DTM and I remade the process.

Path 3d Length is now 11.05Km (similar to the one from my friend...)

This last length should be the most correct one, assuming a forerunner GPS x/y accuracy around 2/3 meters and a true elevation (DTM).

Why does forerunner gives me 7.99 Km instead of 10.1Km? It's 2Km shorter than the one calculated by GPX original data...

And whorse 7.99 vs 11.05Km, in this case we can't compare GPS Z accuracy VS High definition DTM...

I'm not happy with Forerunner distance measurements, not at all.

Cheers
Pietro

  • I can't comment on your distance difference, but GPS smart is not every second. My experience is that precision improve significant by changing to every second and that is what I am using. But hardly enough to explain the difference.

  • As I wrote GPS smart is turned off.. And I set to record each second. 

    So every second forerunner should take a point

  • Not sure if this helps, but Garmin watches don't support 3D distance, except for "outdoor watches" like Fenix (and its cousins, 935/945):

    support.garmin.com/.../

    support.garmin.com/.../

    www8.garmin.com/.../GUID-69434E34-3AD0-41FA-B83B-ECDB462B72DC.html

  • Thank you
    That makes sense, may be a barometer is needed to calculate 3D Distance direct on the watch

    But

    As Forerunner 245 loads data on Garmin site (Connect) there is an elevation correction available, what is it for??

    I made a comparison between real elevation (DTM) and the one downloaded as gpx from connect. There aren't big differences..

    Then I made just now a new comparison

    With the GPX downloaded from connect (7.99Km) I imported the track (that is a 3D track) into Postgis (spatial database) with a metric projection (EPSG 32633).

    Then I used two spatial funcions ST_Length2D(geom) that returns 2D length of a linestring vector feature and the ST_3DLength(geom) function that calculates 3D distance of a 3D vector linear feature

    The result is

    Garmin Connect 7.99Km

    ST_3DLength 10.05 Km (similar -50m- to 10.1Km calculated with QGis - same EPSG see above)

    ST_Length2D 9.84Km

    So, between 7.99Km and 9.84Km (assumed that these are both 2D distances) there is almost 2 Km difference and 2Km difference on 7.99 is 25% error - awful!

    And of course I set to record a point each second on my watch...

  • Another test last week

    I went for a sail trip/contest, so my path is completely flat

    Forerunner measured 13.37nm (24.8Km)

    GIS metric conversion and length measurement is 29Km

    Almost 4Km difference on a 29Km path (14% difference....)

  • That sounds very strange. I can understand that you are not amused.

    Of course I don’t know what is wrong with your watch, maybe a hardware defect. I can only comment on the accuracy of my FR245. I have used it for quite some time now with a couple of firmwares and GPS versions. Activity is running from Garmin with GPS & Glonass (default), with Smart recording (default).

    There is one track, I run often, eventually once a week. It is in on a road in a wood, so some interference from trees. Measured with Google Maps, it is 10,3 km long. My FR245 has given me distances between 10,28 and 10,38 km for this track.

    So all under 1%, mostly even less then 0,5%.

    Do you use a standard activity for recording or something from the move IQ store? I have discussed a topic here in the forums, where a third party activity was showing wrong values, while the Garmin activity was correct.

  • Hi

    I use standard garmin activity. 

    I wrote to garmin support and they told me to flash latest firmware (9.6) and do a full format

    We will see what happens.. 

    Thanks

  • Good luck.

    Maybe it is a stupid question, but did you have a full lock on GPS when you started your activities? (FR 245 confirms that with „green bar“ on the top, searching usually takes 20 seconds up to a minute)

    The only time I had such big differences in measured distance like you described, it was when I ran with a borrowed Polar device and did not wait for the GPS lock at start of the activity, because I was not familiar with it. Ran the 10,3 km and the watch said something like 7 or 8…

  • Yes I always wait for the GPS fix, green bar and beep.. 

    Thnx