Who's right..Google or Garmin?
rouleur23
Posts: 175
Hi
I have a few routes I do on a reg' basis but there is always a big difference in the meters climbed. Example..I have a 52km route that records 470m of ascent on the Garmin yet if I use a site like RidewithGPS (who use Google right?) it come in at 699m. That's a big difference. Who is more likely to be right?
Thanks
Pascal
www.famouscyclingcols.com ride one of the toughest sportives in the biz!
I have a few routes I do on a reg' basis but there is always a big difference in the meters climbed. Example..I have a 52km route that records 470m of ascent on the Garmin yet if I use a site like RidewithGPS (who use Google right?) it come in at 699m. That's a big difference. Who is more likely to be right?
Thanks
Pascal
www.famouscyclingcols.com ride one of the toughest sportives in the biz!
0
Comments
-
It's not an Edge 200 is it? When I upgraded to a 510 the amount I climbed jumped by around a third and now agrees more with the maps.
Then again my 510 gives around 10% more than my wife's 810 when we ride together!0 -
I wouldn't necessarily trust either of them.0
-
On both Garmin Connect and Strava I get slightly different elevation readings.Both have "elevation correction" and I still get different readings.
I recently plotted a route I'd just ridden through Memory map and the elevation was over 1000ft higher compared to both Garmin and Strava.
I personally go with Garmin but that doesn't mean it's right!0 -
Plot the route on one of the cycle planning websites and look at the elevation profile. See which was closest to this. Alternatively learn how to plot an elevation by hand from a paper map if it bothers you a lot.
Or choose one and stick with it. Then if it is wrong it is always wrong by about the same amount. Just an idea.0 -
On both Garmin Connect and Strava I get slightly different elevation readings.Both have "elevation correction" and I still get different readings.
I recently plotted a route I'd just ridden through Memory map and the elevation was over 1000ft higher compared to both Garmin and Strava.
I personally go with Garmin but that doesn't mean it's right!
The trick with using elevation correction is to ensure that you set a known elevation point close to the start of the ride as a calibration point, Garmin then calculates elevation relative to that. I set a point just outside my house (using elevation data off some website) and have found that the Strava elevation correction works great and always returns a profile akin to that ridden.0 -
Interesting!On both Garmin Connect and Strava I get slightly different elevation readings.Both have "elevation correction" and I still get different readings.
I recently plotted a route I'd just ridden through Memory map and the elevation was over 1000ft higher compared to both Garmin and Strava.
I personally go with Garmin but that doesn't mean it's right!
The trick with using elevation correction is to ensure that you set a known elevation point close to the start of the ride as a calibration point, Garmin then calculates elevation relative to that. I set a point just outside my house (using elevation data off some website) and have found that the Strava elevation correction works great and always returns a profile akin to that ridden.
I spoke to someone recently who stated that Garmin etc takes elevation data from Barometric pressure(?) and Memory Map/Anquet/any digital mapping calculates data according to the contour lines crossed?0 -
I think thats correct - so if the barometric pressure changes as you're out cycling - then the altitudes will go wonky.
There should be an option to calibrate the altitude of your home location - so you look that up on an OS map and tell the garmin the altitude so that that part is accurate to set off from.
The google and online mapping functions can only go from what is online - and depends on how accurately your route has been measured.0 -
Interesting!On both Garmin Connect and Strava I get slightly different elevation readings.Both have "elevation correction" and I still get different readings.
I recently plotted a route I'd just ridden through Memory map and the elevation was over 1000ft higher compared to both Garmin and Strava.
I personally go with Garmin but that doesn't mean it's right!
The trick with using elevation correction is to ensure that you set a known elevation point close to the start of the ride as a calibration point, Garmin then calculates elevation relative to that. I set a point just outside my house (using elevation data off some website) and have found that the Strava elevation correction works great and always returns a profile akin to that ridden.
I spoke to someone recently who stated that Garmin etc takes elevation data from Barometric pressure(?) and Memory Map/Anquet/any digital mapping calculates data according to the contour lines crossed?
Should probably clarify what I meant, Strava/Garmin will allow Elevation correction from a "trusted device", such as a 500, 800, etc. These devices work out elevation by changes in barometric pressure so, all the user has to do, is set the start elevation and let the device do the rest.
The Garmin/Strava sites by default determine elevation from tracking known elevations against areas but these can vary greatly in accuracy and range. Hence, the more accurate way is always by device...but with the calibration sorted to assist. The more calibration points the better but who wants that fuss. I know a hilly route because it hurts :-)0 -
-
For this year's Etape, my Garmin 800 measured the elevation at 4,386m.
Garmin Connect's elevation correction puts the total up to 6,327m.
I know that Strava's correction put it even higher, but once c'corrected' I don't think you can change back in Strava.
The official figure for the day was 4,600m.
Personally I'd rather believe the true figure is somewhere between the figure recorded by my Garmin and the official figure.
I've also plotted routes on Strava and the total altitude given has greatly exceeded the reality.
The barometric readings from the Garmin devices is far more accurate than those given by the mapping sites, though still subject to error if atmospheric conditions suddenly change during a ride.0 -
How do you set the start elevation?
On the unit or on Garmin Connect?
https://support.garmin.com/support/searchSupport/case.faces?caseId=%7B6f9f9310-dbe8-11e1-dcef-000000000000%7D0 -
How do you set the start elevation?
On the unit or on Garmin Connect?
https://support.garmin.com/support/searchSupport/case.faces?caseId=%7B6f9f9310-dbe8-11e1-dcef-000000000000%7D
Cheers for that!
I've always been lead to believe that Digital mapping,such as Memory Map,is accurate as it uses Ordnance Survey Maps?The barometric readings from the Garmin devices is far more accurate than those given by the mapping sites, though still subject to error if atmospheric conditions suddenly change during a ride.
I did a Sportive earlier this year and the organisers claim 5,500ft of ascent.Strava came in at 5.350ft.Garmin was 5909ft(5320ft corrected) and Memory Map was 6500ft!
I'm in the Lakes all next week I'm going to set an elevation point then run the routes through Memory map to see the difference!0 -
I've always been lead to believe that Digital mapping,such as Memory Map,is accurate as it uses Ordnance Survey Maps?The barometric readings from the Garmin devices is far more accurate than those given by the mapping sites, though still subject to error if atmospheric conditions suddenly change during a ride.
I did a Sportive earlier this year and the organisers claim 5,500ft of ascent.Strava came in at 5.350ft.Garmin was 5909ft(5320ft corrected) and Memory Map was 6500ft!
I'm in the Lakes all next week I'm going to set an elevation point then run the routes through Memory map to see the difference![/quote]
The Garmin will pick up small rises and falls in the road between the contour lines that a map uses, so unless a really massive scale map was being used, the mapping software will struggle to be as accurate.0 -
I think thats correct - so if the barometric pressure changes as you're out cycling - then the altitudes will go wonky.
Such as this ride (start and finish were both not far from sea level...):
https://connect.garmin.com/modern/activity/765310514
Low pressure system came through while we were touring, it thinks we steadily rose all day when in fact it was mainly coastal.
Or this one, I think some water must have got in the barometer, the start and finish were at the same place. Also got horrendous GPS drift on the descent, the ascent and descent were on the same road:
https://connect.garmin.com/modern/activity/8759272400 -
I've given up on elevation, you can go on a ride where various models of Garmin are being used and the ascent can varey by up to a third. I rode over Weaver Hill close to where I live twice last week, on the first occasion it was 1150' and on the second occasion it was 1000'.0
-
It all depends on the resolution of the base data's DEM (digital elevation model). If the resolution is say 50m then the height value for each cell will be averaged.... Multiply that along a 50mile route then it can result in a significant difference between modelled (strava adjusted) and reality.
Somewhere like the UK has very detailed commercial DEMs available. Places like the middle of a mountain range (little commercial value) will have a more course DEM.0 -
With digital data. I always wondered if errors were created because the software only measures increments of say 10 feet? For example, If you climb 0 to 9 ft it won't register. But when you go past the 10ft mark, the software will register a count. Likewise if you're on rolling hills, going from 9ft upto 11ft back down to 9ft, everytime you pass 10ft mark is it registering another 10ft climbed rather than 1 or 2 ft? Basically the digital resolution only works in increments of 10."The Prince of Wales is now the King of France" - Calton Kirby0
-
Strava don't use them as the cost would be too high. They use the free datasets available which are not as detailed as the Ordnance Survey datasets.Somewhere like the UK has very detailed commercial DEMs available. Places like the middle of a mountain range (little commercial value) will have a more course DEM.
That will depend on how a particular system works. It could be programmed either way.With digital data. I always wondered if errors were created because the software only measures increments of say 10 feet? For example, If you climb 0 to 9 ft it won't register. But when you go past the 10ft mark, the software will register a count. Likewise if you're on rolling hills, going from 9ft upto 11ft back down to 9ft, everytime you pass 10ft mark is it registering another 10ft climbed rather than 1 or 2 ft? Basically the digital resolution only works in increments of 10.0 -
It would be interesting to know what resolution Strava use for their elevation data. Can totally understand the mismatch if its a low res grid.
I'm unsure what data smoothing algorithms Garmin use, but if the head unit collects every few seconds it will plot all small scale increase/decrease in elevation. This will generate higher totals compared to low res DEMs with greater averaged and smoothed surfaces.0 -
I tend to trust RidewithGPS in general terms, but it does have inaccuracies in mountainous areas, as it doesn't understand tunnels that pop under the top parts of cols, and sometimes goes lumpy-bumpy on hairpins and steep sideways slopes. If you zoom in on parts of the actually smooth Col de Rousset, you'll see what I mean. http://ridewithgps.com/routes/102987360
-
Strava state what they use here: https://strava.zendesk.com/entries/20965883-Elevation-for-Your-ActivityIt would be interesting to know what resolution Strava use for their elevation data. Can totally understand the mismatch if its a low res grid.For activities located outside of the US, we consult the ASTER and SRTM databases which have a resolution of 30 meters and 90 meters respectively.
Other route planning sites will most likely be using the same data so will have exactly the same issues as Strava.0 -
I tend to trust RidewithGPS in general terms, but it does have inaccuracies in mountainous areas, as it doesn't understand tunnels that pop under the top parts of cols, and sometimes goes lumpy-bumpy on hairpins and steep sideways slopes. If you zoom in on parts of the actually smooth Col de Rousset, you'll see what I mean. http://ridewithgps.com/routes/10298736
Yes, also roads cut into cliff faces. Routes in the Vercors have a few of these gorges and roads with tunnels. It's a bit disconcerting plotting a route and seeing a 150m climb at 45%.0 -
An example - plot a route from the entrance of the Mont Blanc tunnel to the exit. 11.8km, over 1700m 'climbing'.0
-
To set the elevation on any Garmin with a barometer you can waypoint the start of the ride (front of the house in my case) and edit the altitude saved in the waypont to what you know it to be from OS maps, etc. So when you press the start button the altitude is set accurately at the start of every ride in the future. In Spain I often waypoint the Hotel entrance for example.
If you don't do this you should switch the unit on and leave it stationary for a few minutes so it can triangulate the altitude from the satellites, it can't do this when your moving, If you watch the altitude reading you can see it slowly changing. You can press start when it settles down, after your pre-ride espresso! It then uses the barometer to record altitude on the ride itself. On the old 705 this was all you could do, on the 800, 810, etc, there is a menu option to set the altitude yourself if you know what it is.
I use Strava to plan rides and I find in the UK my height gain recorded by my 810 is mostly close enough to what Strava predicted. On the continent it's very different, Strava estimates are wildly inflated. On a recent Raid Pyrenees the difference was often more than 1000m. Maybe they use a coarser model to the UK?
Strava corrections are meant to be used when your barometric Garmin messed up for some reason, all it does is throw away the barometric data and replace it with its mapping database data.
If you use a GPS without a barometer then the altitude recorded is meaningless and Strava will always use its database.http://www.fachwen.org
https://www.strava.com/athletes/303457
Please note: I’ll no longer engage deeply with anonymous forum users
0 -
Yes the data in the alps is way out - for the Etape the official elevation was 4,600 m, my Garmin read it at 4,400 m and corrected elevations were over 7,000 m :shock:0
-
Yes the data in the alps is way out - for the Etape the official elevation was 4,600 m, my Garmin read it at 4,400 m and corrected elevations were over 7,000 m :shock:
Yes, it's a real nuisance for planning new rides abroad!! Though on our Raid trip the 'team leader' had used Ride with GPS to do the planning and the predicted climbing was a bit more reasonable and not that much different to what we recorded out on the rides. Strava was estimating a lot more and anyone who used the correction button on the website after the ride were massaging their egos!!http://www.fachwen.org
https://www.strava.com/athletes/303457
Please note: I’ll no longer engage deeply with anonymous forum users0 -
Overall I don't think that longish rides in mountain areas will be too wildly out, as long as every road isn't a balcony or going through tunnels. If you look at the profile, you'll probably be able to see the wobbly bits. To an extent I'd guess that coarse elevation data resolution is somewhat less likely to under-report elevation gain in Alpine-type areas as roads tend to be less lumpy-bumpy. So although this recent ride of mine omits the Col de Menée tunnel twice, if you ignore that, the rest of the elevation is about right, adding up the ascending and descending from altitude data from maps - not hard to work out, as the road is either going up or going down in reasonably smooth lines. It would be much harder to work out the accuracy in the same way in Devon though, as the roads are all lumpy-bumpy.Yes the data in the alps is way out - for the Etape the official elevation was 4,600 m, my Garmin read it at 4,400 m and corrected elevations were over 7,000 m :shock:
Yes, it's a real nuisance for planning new rides abroad!! Though on our Raid trip the 'team leader' had used Ride with GPS to do the planning and the predicted climbing was a bit more reasonable and not that much different to what we recorded out on the rides. Strava was estimating a lot more and anyone who used the correction button on the website after the ride were massaging their egos!!0
