Choosing GPS files

[WGroleau]

A couple of times, I have downloaded GPS tracks to load into a mapping/tracking app, and found them very inaccurate. The same thing has happened with downloading maps. Usually, they are close enough to the actual trail or street to be usable, but it would really be nice to get one that's accurate. Is there any place that has "rated" such files? Or can anyone identify one they've found that pretty much follows the trail. (The last one I looked at had a couple of points per kilometer within fifty meters of the trail and straight lines connecting them.)

 

[dougfitz]

There are so many variables that might result in a discrepancy that it would be almost impossible to say whether you were just getting normal variation of the GPS signal or a real problem. Remember that typically handheld GPS will give an accuracy around 10m, irrespective of any cute counters that tell you that they are achieving greater accuracy.

You might also want to check that you are using the correct map datum and spheroid for the mapping data you have loaded. While the WGS84 datum and spheroid are commonly used internationally, Spain, for example is also covered by the European 1979. In Australia, getting the datum wrong can give location errors up to 200m.

Having said this, if the person who has provided the track information you are using hasn't also provided information about the datum that was set when it was recorded, it will be impossible to tell what setting to use. I set the datum to WGS84 in that case, and then do some location checks to see whether that is giving reasonable results.

One way to check the accuracy is to use local oval with a running track, and do several laps on the inside lane. This will allow you to check both distance measuring accuracy and location accuracy.

Finally, there will always be issues about relying on someone else's GPS track. I know, for example, that there were places where I did not walk on the marked path for a variety of reasons. Similarly, many tracks seem to be provided by cyclists, who will not always follow the same path as walkers.

 

[Pilgrim-S-414]

I've used maps.me app on both Caminos with custom gps tracks and found it quite useful. Especially on the Portuguese route which was less well marked than Camino Frances.

 

[kmrice]

If the track consists of a point or two per kilometer, with straight lines between points, it is a track someone plotted on a map, not a track recorded by someone walking it and using a GPS unit to record it. I would try not to use this sort of track; actual conditions on the ground may be very different than the plotter realized.

A recorded track will show the actual path walked or biked by someone. The ones I record sometimes look like I was drunk; they weave a bit. This is because they follow my actual walk, dodging small obstacles, avoiding puddles or whatever. They also show minor deviations for when nature calls. Such a track looks very different than a plotted one; the difference is quite obvious.

A recorded track can vary from the actual path as GPS is not perfectly accurate. In the distances covered by a walker in a day, however, it will usually be consistently inaccurate, though, so if it is 20 feet off the trail to the north, you can just keep 20 feet to the south all day and you will be fine.

I don't think you mentioned which Camino you are looking at, but the paths of the Camino Frances and the Le Puy Camino are very well marked, usually heavily used, and quite visible; I only check my GPS occasionally to confirm I am on the path, or to figure out which way to turn at an intersection if the turn is not marked (almost all are).

There are a number of sites with recorded tracks for many of the Camino routes. I think I downloaded mine from www.goldesalco.com.

Good luck and buen Camino.

 

[peregrina2000]

First of all, I am a GPS convert but a very low tech person. I would NEVER use a GPS on the Camino Francés. (Sorry, I know that's not what you asked.) I use a GPS on solitary caminos, but I have worked out a deal with myself to avoid being one of those people who walks with her head pointed downward to the GPS. When Im using my GPS, I pull it out of the side of my pack only if I have gone for 15 minutes without an arrow or if I am at an intersection where there is no marking and there are several ways to go. That will never happen on the Camino Francés, Primitivo, Norte, Invierno, Vdlp. So my unsolicited opinion is that they are a distraction on those caminos. If you're walking the Olvidado, Catalán, Vadiniense, Castellano-Aragonés, Ebro, etc, then I would download GPS tracks from wikiloc.com. I have had great luck on that site.

Make sure to download single stages, not the whole route, or you will have 500 waypoints spread over a huge distance. When you download from wikiloc, they ask you whether you want the 500 point version or the original version (that is because most GPS devices will only incorporate 500 points per track, as I understand it). For a 30 km stage, the average point number is way over 500, so I imagine if you download a whole camino in one file, you will get a very imperfect set of points.

 

[dougfitz]

If the track consists of a point or two per kilometer, with straight lines between points, it is a track someone plotted on a map, not a track recorded by someone walking it and using a GPS unit to record it. I would try not to use this sort of track; actual conditions on the ground may be very different than the plotter realized.

I don't agree with the view that a sparse track is necessarily plotted, rather than recorded. I have attached three .gpx files that, somewhat serendipitously, I was working on for a local walk. The original track (Hall to the Rock #1) is an amalgam of several walks, but it contains nearly 1500 track points, most of which add very little useful detail in terms of providing route guidance. The other two tracks have about 100 and about 40 track points, or an average of five and two point a km for this route. They are both extracts of the original, and are NOT plots done on a mapping application. While the denser of these tracks might be a little more usable, both are perfectly adequate for what might be a complicated route, but in this case on well defined path.

More importantly, in terms of sharing the information, the original file size of 240KB has been reduced successively to 10KB and 6KB, or about 1/25th and 1/40th of the original file size.

I think it is perfectly reasonable to use this sort of track, it is a relatively common way of sharing track information, and I think it incorrect to suggest that it does not reflect conditions on the ground. I suspect the sparse tracks that I do see are because someone has done the sensible thing, and thinned out the track data they have collected to a usable minimum.

A recorded track can vary from the actual path as GPS is not perfectly accurate. In the distances covered by a walker in a day, however, it will usually be consistently inaccurate, though, so if it is 20 feet off the trail to the north, you can just keep 20 feet to the south all day and you will be fine.

I rather suspect a consistent bias would need to be explained by some other bias than the inherent inaccuracy in the GPS signal. One way of demonstrating this is to leave your GPS on at home for a couple of hours. If, as you suggest, there is a consistent bias all the track points would be distributed consistently in a particular direction. My observation is that when I do this, the points are distributed randomly around what I know to be the location of my study.

ps the web address for Godesalco is http://www.godesalco.com/.

 

[kmrice]

dougfitz's point is well taken: a track with few points might be a track someone extracted from a larger file. But again, it might not be. A very dense track, however, will be one which was actually walked or biked by someone, and, at least for me, is more useful and more reliable. I'd always rather have more data, not less.

It is certainly true that the denser the track, the larger the file gets. But, given the amounts of memory available on devices these days, I'd prefer a dense, 240KB, file to a sparser, 6KB, one with only 2 points per kilometer. The very dense tracks I used for the Le Puy Camino last summer took up an insignificant amount of memory on a 64GB phone and were small enough to share easily.