This is the original text; see also first update and latest presentation


Bruxelles, le 01-03-2006

notre ref RCH/50358

à l'attention de Janice Sutherland et Martin H-S

Has anyone ever plotted / recorded the Camino Frances with a GPS Device (Garmin)?

Here is a file StJacques2005byG7ToWin.csv of our track last year. Its a dump of Latitudes, Longitudes, "times", altitudes and "depths". Format is *.csv and so it should be readable in just about any software that you happen to have.

I also made other versions ( *.gpx, *.g7t *.gdn, *.xls and *.xml), but *.csv should be enough.

You can, for example, use a spreadsheet programme to plot a graph of the track (Latitude versus Longitude), or read into GIS plotting software, or make a track and upload to your GPS. Here is a nice link on FR coordinate systems, which may help you get your mind round the various Lambert etc. systems you will find on IGN maps.

There are of course lots of other useful sites

I'd be interested to hear how you get on.


Richard Hardwick


You will find some oddities in StJacques2005byG7ToWin.csv. Some at the start, because I was learning how to use the Etrex as we went along, others due to the well known deficiencies of the Garmin Etrex (see comment by RichardH. here). These deficiences are:-

  1. Indicated "times of observation" are meaningless. This is because when it saves tracks, Etrex silently decides not to keep the times of the observations. It overwrites each individual value by X where X = the value when the track was saved! see e.g. Mapping Hacks page 260

    I think possible solutions are:-

    • Either - Instead of saving tracks each evening, upload to a datalogger each evening, and clear the memory.

    • Or - Chuck the Etrex, buy something with bigger memory, and never save tracks, just let the memory get fuller.

  2. When satellite reception is poor (e.g. during our descent into Conques), there is a crazy zigzagging of around +/- 0.002 deg Lat/deg Long in the recorded track.

    Solutions ?

    • If the times of observation are available, its simple enough to eliminate the false points automatically - calculate distance over time = velocity between points, and reject all values (metres per second) that are impossibly high.

    • If you only have stored tracks, i.e. if times of observation are not available, I dont know of any way of removing the zig-zags. I tried 3-point smoothing and 5-point smoothing, but they don't work well enough to be useful.


Back to homepage
This url is

Original file 50358.html

Version 002

Dated 2006-06-02T17-48-00