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

problems with distance axis using data from 301

Hi,

I'm enjoying using garmin connect, but I'm having some issues with the graphs generated when uploading run data from my garmin 301. Specifically, if I go for a run of (let's say) 8 miles, the activity summary will correctly display the distance, time, and so on, and the "timing" chart for the activity will look fine, but any chart where the x-axis is distance is messed up -- the 8 miles of run data are rendered onto an x-axis of 5 miles, and hovering over lines returns funny values for the data points.

For example, here's a recent activity I uploaded:

http://connect.garmin.com/activity/25529697

Note the distance in the summary field is correct. Now, click on the "elevation" tab, and look at the max value of the x-axis -- it's wrong. Similarly, if you run the player (click on the player button), distances will be messed up.

If you export a TCX file and import it into garmin training center, things look fine there. Similarly, if you upload an activity directly from the device into training center, things look fine there.

Any ideas?

Thanks very much!
--Steve
  • I've had a look at the data using excel. The column that shows the distance covered up to each track point reaches a total of 9,500 metres whereas the total distance according to the total distance column is nearly 13,000 metres. There are a few occaisions where it loses satellite contact and during these times it does not update your position, speed or distance. When it re-aquires you don't appear to have moved much. There is one occaision where it loses it for 40 seconds during which time you only move 10 metres - is this possible ? did you get held up ?
    It must be calculating the distance using two different methods and the accumulating version is undermeasuring somehow.
  • Thanks for the detailed reply, twostage, and for the help investigating!

    No, I was never held up. I was running on the burke gilman trail in Seattle, and ran continuously without any stops at all -- I literally didn't stop moving for the full run. The trail does go through some areas where neighboring trees and hills caused my 301 to lose satellite coverage, but I was running between 7 and 8.5 mph pretty much the full time.

    Looking through the data file (good idea!), it looks like when satellite coverage drops out (i.e., no lat/long data points in a trackpoint), that's when the distance stays "static", and is not properly updated when satellite coverage resumes.

    Garmin training center must be doing something to compensate for it -- maybe it's computing distance/speed/pace using the lat/long coordinates directly, rather than the reported "distancemeters" data...?

    Thanks!
  • So, here's some more intrigue.

    • When I import data directly from the 301 into garmin training center, the distances look fine.
    • When I import data directly from the 301 into garmin connect (using the web browser plugin), the distances are broken.
    • When I export a .tcx file from garmin connect and import that into garmin training center, the distances are broken as well.
    • But, when I export a .gpx file from garmin connect and import into garmin training center, the distances are correct! [unfortunately, even though the .gpx file contains heartrate data, garmin connect doesn't seem to know how to deal with it when importing from a .gpx file, so just using .gpx doesn't solve my problem. there are also some discrepancies in the total elevation gain when using .gpx.]
    • Similarly, when I export a .gpx file from garmin connect, and re-import it into garmin connect, the distance stuff is correct!!!


    So, my guess is that whatever is converting from the raw gps tracks of the 301 into .tcx format (either the plugin on my client or garmin connect itself) is making mistakes and not accumulating distance correctly after a satellite dropout. The .gpx format doesn't include accumulated distance, so whatever is calculating it from .gpx tracks (either in garmin training center or in garmin connect) isn't making the mistake.

    I'll try updating my 301's firmware (i'm using 3.0, looks like 3.2 is available).

    If that doesn't work, I bet I can write a quick n' dirty script to fix the distance calculations in the .tcx file, though that will send me down the path of figuring out how to compensate for gps jitter errors. :(
  • The differences between what works as far as the distance is concerned through the various exports and imports must depend on which fields are being used for distance calculation. It would be interesting to see the raw data as it comes out the watch rather than the tcx file which has been processed. Is that possible ?

    I think the firmware upgrade should help.