Discussion
Looking at my Endomondo logs, the blue trace of my route appears to wander off the actual satellite image of the route I have taken some 10m here and there. I'm using my phone which is stuffed into my saddlebag next to a pair of Magicshine batteries.
Is it likely that my trace would improve if I gave the phone better "visibility" of the sky, or is this accuracy a consequence of using a phone (as opposed to a dedicated GPS device) or some slight error in the overlaying of the app's data onto Google maps?
I'm considering using one of those waterproof arm-bands for my phone, to give it a better chance. Worth a shot or not going to give the desired results?
Is it likely that my trace would improve if I gave the phone better "visibility" of the sky, or is this accuracy a consequence of using a phone (as opposed to a dedicated GPS device) or some slight error in the overlaying of the app's data onto Google maps?
I'm considering using one of those waterproof arm-bands for my phone, to give it a better chance. Worth a shot or not going to give the desired results?
Think the handle bar mounted GPS devices tend to be more accurate. Partly due to the better visibility and partly due to them polling the satellites more often.
Saying that my Iphone 4s used to be pretty reliable. I always carried it in my back pocket of a jersey though.
You'd assume that batteries, a seat and your ass would restrict the signal somewhat but I couldn't say for defo that it does.
Saying that my Iphone 4s used to be pretty reliable. I always carried it in my back pocket of a jersey though.
You'd assume that batteries, a seat and your ass would restrict the signal somewhat but I couldn't say for defo that it does.
dedicated GPS unit poll the satelites more often than phone apps, giving greater accuracy which shows up as quite a lot of distance on a long ride, I ran both my Garmin and my android phone on a 100mi sportive and the phone added about 2mi to the distance due to the zigzag nature of it's plotting.
They are also notorious for giving false times on short Strava segmants for the same reason, the gap between GPS polling info and the way Strava handles the data leads to it 'best guessing' rider position at any given moment.
iirc last gen garmin, 200, 500, 800 was 1 second polling and android is 5, could be wrong will look it up.
They are also notorious for giving false times on short Strava segmants for the same reason, the gap between GPS polling info and the way Strava handles the data leads to it 'best guessing' rider position at any given moment.
iirc last gen garmin, 200, 500, 800 was 1 second polling and android is 5, could be wrong will look it up.
Garmin Edge500 here. It can vary significantly on three consecutive laps of a local lake, yet the route is no more than 4 to 6 metres wide the whole way around. It places some laps off laterally into the trees, yet at other times it'll overlay the laps virtually on top of one another, perfectly on the satellite image of the track. I've no idea why, although I do notice that the mapping data for roads etc is sometimes not exactly over the actual road/track on satellite photos, and the image changes from satellite photo to aerial photography images when you zoom in tight.
As an aside, I have found that it 'drops out' less when in the out-front mount, horizontal, on my road bike, when compared to the riser stem and 'standard' mount on my MTB which lean it back at an angle. This is exacerbated by the angle of ascent on hills, resulting in 'holes' in the data when the GPS loses satellite connection when I'm out on my MTB. Another 'out-front' mount is what is needed to solve that particular issue (for me, anyway).
As an aside, I have found that it 'drops out' less when in the out-front mount, horizontal, on my road bike, when compared to the riser stem and 'standard' mount on my MTB which lean it back at an angle. This is exacerbated by the angle of ascent on hills, resulting in 'holes' in the data when the GPS loses satellite connection when I'm out on my MTB. Another 'out-front' mount is what is needed to solve that particular issue (for me, anyway).
Jayfish said:
dedicated GPS unit poll the satelites more often than phone apps, giving greater accuracy which shows up as quite a lot of distance on a long ride, I ran both my Garmin and my android phone on a 100mi sportive and the phone added about 2mi to the distance due to the zigzag nature of it's plotting.
They are also notorious for giving false times on short Strava segmants for the same reason, the gap between GPS polling info and the way Strava handles the data leads to it 'best guessing' rider position at any given moment.
iirc last gen garmin, 200, 500, 800 was 1 second polling and android is 5, could be wrong will look it up.
There's no set polling rates for mobile phones, it's just when they sense the change in direction that it will poll. They are also notorious for giving false times on short Strava segmants for the same reason, the gap between GPS polling info and the way Strava handles the data leads to it 'best guessing' rider position at any given moment.
iirc last gen garmin, 200, 500, 800 was 1 second polling and android is 5, could be wrong will look it up.
You can fix GPS traces uploaded to Strava here, it definitely works for roads, dunno about offroad
http://strava-tools.raceshape.com/snap/
Jayfish said:
iirc last gen garmin, 200, 500, 800 was 1 second polling and android is 5, could be wrong will look it up.
I just had a look at the GPX file from a Strava Android app plot, and it looks like it's polling every 1-2 seconds. The timestamps are mostly 1 second apart, but sometimes the interval is 2 seconds. So not bad really...I've always found that the phone data is pretty accurate, to the point of showing which side of the road I'm riding on, and the willy I drew in a large car park the other day showed up nicely as well
There are a number of factors in GPS accuracy. The receivers in phones vary in quality quite a bit, so some will be a lot better than others. The atmospheric conditions make a big difference, and so does the particular position of the satellite constellation. Phones which are capable of receiving the GLONASS and Beidou systems have access to more satellites and so aren't so affected by that. Giving the receiver a better view of the sky won't hurt though.
One thing that phones don't have at the moment, that dedicated GPS receivers usually do is wide area augmentation. This uses a geostationary satellite to transmit corrections for current atmospheric conditions and can improve the reception to 2-3m accuracy. In Europe it is called EGNOS, and WAAS in the US. I don't know if the GPS based cycle computers use WAAS enabled receivers, but they should list it in the specs if they do.
One thing that phones don't have at the moment, that dedicated GPS receivers usually do is wide area augmentation. This uses a geostationary satellite to transmit corrections for current atmospheric conditions and can improve the reception to 2-3m accuracy. In Europe it is called EGNOS, and WAAS in the US. I don't know if the GPS based cycle computers use WAAS enabled receivers, but they should list it in the specs if they do.
Gassing Station | Pedal Powered | Top of Page | What's New | My Stuff