Roger Longbottom
Guest
W
Last edited:
W
- Thread starter Roger Longbottom
- Start date
-
- Tags
- strava
Page may contain affiliate links. Please see terms for details.
cyberknight
As long as I breathe, I attack.
- Location
- Land of confusion
thats normal
strava auto corrects your data based on better maps and elevation data.
strava auto corrects your data based on better maps and elevation data.
steveindenmark
Legendary Member
Which bike computer are you using?
HLaB
Marie Attoinette Fan
Strava has a higher threshold for moving speed and your computer will detect a stop faster. 0.2 of a difference is normal, if you go out on a ride with zero stops the computer and strava should be closer.
https://support.strava.com/hc/en-us/articles/115001188684-Moving-Time-Speed-and-Pace-Calculations
https://support.strava.com/hc/en-us/articles/115001188684-Moving-Time-Speed-and-Pace-Calculations
Dogtrousers
Kilometre nibbler
If you calculate something that depends on an estimation algorithm - like moving average speed, total elevation gain, watts or calories then you will get different answers depending on who does the calculating : different websites, GPS devices etc.
That's why you should choose one method and stick with it. Never compare between methods. And don't ask what is the most accurate. Just pick one and stick with it.
That's why you should choose one method and stick with it. Never compare between methods. And don't ask what is the most accurate. Just pick one and stick with it.
Dogtrousers
Kilometre nibbler
No, it will upload the ride as a series of location points with the time at which the location was recorded. Strava will interpret this string of points as a track, and from the position and time will infer whether the GPS was in motion.
Your GPS will also, independently figure out whether or not it is in motion and will come up with a moving time.
The two will differ.
Dogtrousers
Kilometre nibbler
They're different tracks. It's fairly tricky to detect on a GPS track whether the recorder is stopped. Even when stopped the points will dither around. So two different tracks will give different results.
Seriously, choose one method to trust and stick with it. Ignore other methods. Comparing methods will just cause premature ageing and general woe.
HLaB
Marie Attoinette Fan
Its quite simple, the bike computer autopauses at say 4mph but strava doesn't autopause to say 2mph (I dont know the exact mph numbers) so if you're travelling at 3mph for a second or two strava hasn't autopaused but the computer has paused the timer. Over the course of a ride these differences will easily add up to 40sec. 0.2mph is very normal nearly all my rides are 0.1-0.2mph slower in strava than on the computer unless its a ride with minimal stops.
figbat
Slippery scientist
- Location
- South Oxfordshire, UK
When the data goes to Strava, it goes as raw waypoints, it doesn't go as the 'processed' speed and distance data. Strava then treats that data as if it had recorded it itself, interpreting the waypoints and processing it in its own way to give speed and distance results. Your first ride coincidentally gave the same time result but this is likely to be the exception rather than norm.
Drago
Legendary Member
- Location
- Suburban Poshshire
A conventional vehicle speedo calculates velocity along a surface. That surface undulates with the terrain, and turns left and right with the curves.
GPS uses multiple satellites to calculate your position in 3D space. It may "see" a dozen satellites, but the algorithm uses only 2 at a time to make the calculation. GPS then recalculates it very quickly, dozens of times a second, and the more satellites it can see the more it can switch to different birds to make the next calculation, and the more accurate it becomes.
now, this is where it gets sticky. GPS calculates a position in 3D space. Your devices then calculates the next position in 3D space, and then makes a back calculation using the distance covered divided by time to get a velocity expressed as a straight line in 3D space between those 2 points. Unfortunately, it calculates that velocity as a straight line between 2 points covered in x time, when in reality the terrain is not a dead flat plane between 2 points in 3D space, and you are not travelling in a dead straight line in 3D space. The road has bumps, hills, dips, twists, curves, so the actual distance you have travelled along a surface will differ from the distance you have travelled if measured in a straight line in space between any 2 given points on that journey. The faster you go, the more windy or hilly the road, the greater potential cumulative discrepancy.
What Strava can do after a ride is take the GPS data, overlay it against the terrain/elevation data, and make corrections to bring it closer to the true ground speed/distance, rather than thousands of discrete straight line calculations. It isn't perfect, but gets a closer result to true ground speed/distance as measured by direct means.
To really confound things, velocity has to be accounted for. Each GPS calculation gives a number, and by adding all these calculations up it gives you your speed. In reality, your speed isn't constant, even for a split second. Draw your speed on a graph and it'll be a smooth up and down line, varying with velocity - draw it using GPS data and it'll be lots of dots on the graph that looks like a smooth line, but zoom in and you'll see straight lines connecting each dot. Think of it as akin to analogue and digital music reproduction.
I hope that makes sense! Point being, there will be minor differences, and that difference can change dependent upon velocity and elevation change - the difference will not remain as a fixed percentage, but will vary due to the factors I've (badly) described above. As mentioned above, it happens and there is nothing you can do beyond selecting a preferred method of recording and sticking with it.
GPS uses multiple satellites to calculate your position in 3D space. It may "see" a dozen satellites, but the algorithm uses only 2 at a time to make the calculation. GPS then recalculates it very quickly, dozens of times a second, and the more satellites it can see the more it can switch to different birds to make the next calculation, and the more accurate it becomes.
now, this is where it gets sticky. GPS calculates a position in 3D space. Your devices then calculates the next position in 3D space, and then makes a back calculation using the distance covered divided by time to get a velocity expressed as a straight line in 3D space between those 2 points. Unfortunately, it calculates that velocity as a straight line between 2 points covered in x time, when in reality the terrain is not a dead flat plane between 2 points in 3D space, and you are not travelling in a dead straight line in 3D space. The road has bumps, hills, dips, twists, curves, so the actual distance you have travelled along a surface will differ from the distance you have travelled if measured in a straight line in space between any 2 given points on that journey. The faster you go, the more windy or hilly the road, the greater potential cumulative discrepancy.
What Strava can do after a ride is take the GPS data, overlay it against the terrain/elevation data, and make corrections to bring it closer to the true ground speed/distance, rather than thousands of discrete straight line calculations. It isn't perfect, but gets a closer result to true ground speed/distance as measured by direct means.
To really confound things, velocity has to be accounted for. Each GPS calculation gives a number, and by adding all these calculations up it gives you your speed. In reality, your speed isn't constant, even for a split second. Draw your speed on a graph and it'll be a smooth up and down line, varying with velocity - draw it using GPS data and it'll be lots of dots on the graph that looks like a smooth line, but zoom in and you'll see straight lines connecting each dot. Think of it as akin to analogue and digital music reproduction.
I hope that makes sense! Point being, there will be minor differences, and that difference can change dependent upon velocity and elevation change - the difference will not remain as a fixed percentage, but will vary due to the factors I've (badly) described above. As mentioned above, it happens and there is nothing you can do beyond selecting a preferred method of recording and sticking with it.
Last edited:
Dogtrousers
Kilometre nibbler
Not necessarily all that puzzling. The computer has written a gpx (or whatever) file for a particular stretch of time. It's also displaying an elapsed time, but these could well be different.
Website analysis of GPX files and instantaneous readouts of computers are always going to differ. Also different websites will disagree about things like average speed or elevation.
Just don't compare different methods. Choose one as your "gold standard" and stick with it.
I have a hideous monster excel file I have built over the years that pulls all kinds of stats and graphs out of my gpx files. Absolute values differ somewhat from those given by Strava or RideWithGPS or my GPS itself. I expect that.
newfhouse
Resolutely on topic
I had always assumed that, providing I haven’t set my Garmin to conserve power and only save according to its own algorithm, each line of the file represents a decode of the relevant NMEA sentence from the GPS engine. As such, the start and finish times, from which elapsed time would presumably be derived, should be accurate to within the sample interval. Having just had a look at a couple of files I see that I am wrong, but not totally.
There is a metadata time header which matches the first saved fix, but no footer, just the last fix. Of the files I have inspected the elapsed time shown on Strava exactly matches the time difference between the first and last fixes in the .gpx.
Every day is a school day, but I’m still puzzled by the OP’s elapsed time difference. I wonder if there is a difference between Garmin and Bryton files?
There is a metadata time header which matches the first saved fix, but no footer, just the last fix. Of the files I have inspected the elapsed time shown on Strava exactly matches the time difference between the first and last fixes in the .gpx.
Every day is a school day, but I’m still puzzled by the OP’s elapsed time difference. I wonder if there is a difference between Garmin and Bryton files?
Dogtrousers
Kilometre nibbler
Not necessarily. All, or most of the Garmins I've used have a "sample interval" (or similar - I forget the name). "Auto" will keep the file size down by, e.g., recording more points on bends and fewer on straights.
The start and finish times of a "trip" for which stats are shown on screen may not match the start and end timed of the saved file, because they are different things.
I never assume that my trip info as shown on the device will match the saved info in the GPX file, because that's just not how things work. For instance when I do my 100km and 100mile challenge rides I always ensure the trip odometer is at least 101 miles or km because the ride may "shrink" when saved to GPX file.
I upload all my stuff to Ridewithgps, and this is what I trust for my stats. I know that the stats will be different to what my GPS device shows, and I know that if I upload the same file to Strava they will be different again. It's the way things are.
