Why such different elevation corrections? (same route logg 1790m, 2615m, 3468m elev gain)

I rode with a friend the other day . I use a Garmin FR 305, he uses an IPhone. 

When I uploaded FR305 -> Strava, the elevation gain is 2,615m, see  http://www.strava.com/activities/138096662

When I uploaded FR305 -> Garmin Connect, the elevation gain is 1.790m (with elevation correction enabled)  and 1.724 m (with elevation correction disabled)  see http://connect.garmin.com/activity/494258370

My friend uploaded from IPhone Strava app , he logged 3,468m, see  http://www.strava.com/activities/138064391 (maybe this activity is not visible to public) 

The Garmin FR 305 is pretty accurate, I have compared data from Garmin Edge devices, and the difference < 5%

It would be good if it was possible to disable the elevation correction in Strava, it is not working very well. Garmin Connect. has the possibility to disable elevation correction even though it does a good job.  Maybe the elevation correction algorithm / database works fine in the US but where I am, it is no good.

31

评论

  • 正式评论

    We've announced some changes to our method of calculating elevation data for Strava Activities. At the end of the article is a feedback form you can fill out to ask questions and report issues. Thanks and let us know what you think. 

    评论操作 固定链接
  • Thanks Gary - I'll try to offer up an explanation, and I hope it helps. 

    A barometric altimeter best measures relative elevation changes based on atmospheric pressure. Since changes in atmosphere pressure are very small, and changes are gradual, the resulting dataset will be well "smoothed". Also, the changes in elevation relative to each other will be accurate, but the absolute readings from your Garmin, for example, may not read the exact elevation at that location. 

    An elevation database created by surveying the land surface will essentially divide the whole world into a grid of squares, with each square reading one value for the average elevation measured in that area. When we use this database to calculate your elevation, we see which "square" each of your GPS coordinates falls into, and use that as the elevation value. A few problems - for much of the world, the dataset is divided into squares measuring 30 meters on each side. For locations with complex topography, a lot can change from one square to another, and the average value of one square might only paint a limited picture of the elevation changes in that square. Pair that with GPS accuracy, and you may have a decent amount of recorded GPS coordinates that place you in a "square" that you were not actually in. Thirdly, this resulting elevation dataset for your activity may not be very smooth, and may have large changes in elevation from one point to another. This can cause issues with inflated elevation readings. 

    Also keep in mind that when you recording with two devices, you're generating two different GPS files, and each GPS will record different coordinates even if they traveled the same route. GPS is a bit of a whacky technology, and has an average of + or - 50 feet accuracy pretty much at all times. So, if you take two different recordings, and correct them both using the elevation database, you will get two different values (also recording intervals play in here, and a Garmin has a variable recording interval while the mobile app is recording every 1-2 seconds). 

    What I'm attempting to accomplish here is to paint a more detailed picture of the factors behind the scenes with your elevation correction, and try to explain some of the subtleties of elevation data, and why there can be some variation based on each GPS recording, device, and method of data collection used. 

    Let me know if that explanation helps! 

    3
    评论操作 固定链接
  • I'd like to send my congrats for Strava team for the great job they have done about elevation corrections here in Rio de Janeiro. Probably, it isn't perfect but the hay done a big huge great improvement. The elevations are now much more realistic. Thank you Guys. Well Done!

    May I ask you if Is it possible to correct the old data?

    Thx

    2
    评论操作 固定链接
  • Doug, I didn’t realize my length of use of Strava has anything to do with the validity or importance of my feedback and usage. I pay for a Premium membership just like anyone that’s been here for longer. And for the record, I’ve been a Strava member for almost 2 years, I think the 6 months shown when you click a person’s name here is for the "Support" side of Strava.

    Good for you in Hobart. Here in Colorado, I’ve already lost 4,000’ in elevation this year across because Strava refuses to allow barometric elevation data from devices that aren’t on their mysterious "approved" list. Instead, Strava is replacing my barometric data with it’s own DEM data, and it’s quite off here in the front range of Colorado. 

    Quite frankly, I don’t think Strava should be so strict in allowing barometric altimeter data. If it’s in your file, it should be utilized. It shouldn’t matter what device recorded the data. At least then it’s up to the user to determine if their device is accurate and if they’d prefer to use Strava’s calculations or not. 

    2
    评论操作 固定链接
  • I'm into a kind of cruzade against Strava forcing elevation correction in devices, other than a few, even if they have barometric altimeter.

    I didn't found until now any cientific explanation that GPS elevation is less precise than barometric for cycling purposes. On the other hand I've seen many times barometric altimeters "pass away" in several atmosferic conditions such as rain.

    So I had the following test: downloading the GPX track from my 12 years old Garmin Etrex Legend and convert it to TCX format. I edited the file and added information saying device used was an Edge 500. Finaly I uploaded the file in Strava web page. This is the result: https://www.strava.com/activities/332985127
    Now tell me if this is less precise than an Edge series device?

    I hope you, Strava guys, don't come with some schema to avoid this trick and, instead, stop impose elevation correction. If you keep this I'm forced to believe in some conspiration theory about Strava only wanting us to spend money in fancy gadgets :)

    1
    评论操作 固定链接
  • I guess this is the problem, the lowpass filter should filter out more and not let through all these small ups and downs, it should probably disregard any altitude difference below 10 meters.

    But again, why not allow the users to disable the altitude correction and use the raw data? I have followed this for more than 2 years, compared data from various devices and frankly  9 times of 10, the raw gps height data is better than the corrected (where I ride) 

    1
    评论操作 固定链接
  • I guess this is the problem, the lowpass filter should filter out more and not let through all these small ups and downs, it should probably disregard any altitude difference below 10 meters. - Jens Kastensson

    That's what it does now.  The problem with that is it ignores reality.  I can do repeats all day on 9.2 meter inclines (which we have plenty of) - the elevation shows correctly in the data but Strava always summarizes it to 0.  The majority of my rides get summed to 0, which is one of the reasons my premium dollars go to Runkeeper.

    1
    评论操作 固定链接
  • That appears to be correct, Liam. 

    1
    评论操作 固定链接
  • Well what I see with my datas and what I read here I think Strava got a serious problem with their own algorithms... not only elevation. 

    Elle Anderson you tell 2 years ago yours and GarminConnects site use same maps and other data for elevation correct? Its still same or some changes was made? 

    Because Its still really different between these two sites. Garmin shows me data as you when I upload first. After correction on both sites shows me huge differents. For example: GarminConnect: 1461m Strava: 1309m. Some times its much more around 500m different is almost rule.

    Why? How? If its still same as you say 2 years ago how is this possible? Cant be same as you see. 


    1
    评论操作 固定链接
  • Charly:

    Yep that is the problem I see and would like to see a change so it will be a fair game for everyone.

    Not sure if some captured elevation numbers are 100% inflated, but my friend with his Garmin 800 had a ride a few months back that got 6,000+ feet captured, and out of curiosity, I remapped his route with ridewithgps and it only got around 3,700+ ft. Assuming ridewithgps is reasonable accurate, that is roughly 38% inflated, which is a lot. Now image if my friend and I try to complete for top spot on the Leaderboard every week on the elevation gain category, assuming we ride together the whole week, he would beat me every time, as my Strava corrected elevation gain would be shaved off significantly.

    The worst part is that Strava is using multiple methods to determine elevation gain data. As an exercise, I mapped a route on Strava's own route building that had around 1,800+ ft. I rode this route and uploaded it back to Strava. To my surprise, Strava activity corrected it to be 1,300+ ft. This shows that the system is pretty much f***up. I believe that Strava used to overestimate elevation data, and people were angry about it, so they decided to underestimate on purpose, but now that affect those of us rely on their correction. I guess that it is hard to make both camps happy.

    Again, I really don't care whatever calculation method being used here, just make it consistent and fair for everyone.

    1
    评论操作 固定链接
  • Hello everyone,

    We are wrapping up an exciting project to address your concerns specifically regarding the accuracy of mobile elevation gain metrics. When we are ready to release more information, I'll be sure to keep this post informed. 

    1
    评论操作 固定链接
  • There is a workaround that involves a bit of manual editing of the GPX file and forces Strava to accept the elevation data from the GPX file (regardless of what device it was recorded with) and not override it with Strava's corrected elevation. 

    To force the workaround you need to download GPX for your activity and edit it in any text editor. Look for the creator field in the beginning of the file that looks like creator="your device name here" and add " with barometer" at the end of the field. After editing it should look like this: creator="your device name here with barometer".

    Once done editing manually upload this GPX and voila! 

    1
    评论操作 固定链接
  • Hi Jens,

    Thanks for passing along the good examples - it is interested to see how other sites handle elevation gain statistics. 

    We use the same database as Garmin for our corrections, so it's unclear why the two differ so drastically. I'm going to pass these examples on to our developers for further research! 

    0
    评论操作 固定链接
  • Hi, I be the same problem. I m using a Samsung S4 whit barometric sensor. The elevation registered during the ride is correct... But when I upload the track it add about 30% elevation. Is it possible to disable correction algorithm?

    0
    评论操作 固定链接
  • I would like to second this same issue thou I am using a Garmin 500 and also recording on Strava iPhone app and have noticed the significant disrepancies.  

    0
    评论操作 固定链接
  • same issue here with Garmin 500 compared with regular major differences with my cycling group

    0
    评论操作 固定链接
  • Hi all

    I have a cycling GPS Bryton 40T  with barometric sensor. The  elevation of all the files I uploaded on Strava (.tcx or .gpx is the same) were corrected in a wrong way (about 30% more). The same file uploaded on the Bryton site give the correct indication

    0
    评论操作 固定链接
  • I have the same problem with elevation. I use android application and when the ride is first uploaded it's very accurate in terms of elevation, but then Strava's elevation correction kicks and messes up the whole thing!! On the last ride at first it said ~540m of elevation which was very accurate but still a bit too much. After the "correction" it jumped to ~740 and it's totally inaccurate. And this is just on a short 44km ride, when there are longer rides it's so much worse. The GPS data from the phone is fine, the elevation correction is absolutely worthless and misleading. Please fix it already or at least add an option to revert the elevation correction.

    0
    评论操作 固定链接
  • I have the same issue. 

    Gained altitule of my workout below was about 3000m, but it was incorrectly corrected.

    The elevation is 6800m, twice as the record!

    It seems to sum up tiny elevational noises.

     

    http://app.strava.com/activities/156559897

     

    0
    评论操作 固定链接
  • Hi Makoto - Thanks for the information. I've sent your file over to the engineering team, and will work on getting this added to our list of devices with "trusted" elevation data. After that happens, Strava will be able to read the elevation data collected by your device. 

    Gio - regarding your Bryton device, I believe we're working to do the same thing if we haven't done so already. 

    0
    评论操作 固定链接
  • I have the same issue. I am getting about a 500 meter difference between iPhone uploads and Garmin 500 uploads, both devices were used at the same time. The delta is larger after the automatic Strava 'elevation' correction is applied. The iPhone gives consistently higher total elevation gains than the Garmin 500.

     

    Thanks

    0
    评论操作 固定链接
  • Perhaps some of you would be interested in adding your votes and comments to this feature suggestion:

    https://strava.zendesk.com/entries/35578550-Revert-to-GPS-based-altitude-when-corrected-elevation-is-erroneous

    Also, there is an interesting discussion on this topic happening near the end of this forum: https://strava.zendesk.com/entries/25860845-Option-to-undo-Elevation-Correction-and-revert-to-Barometric-Altimeter-

    @Gary - your observations are to be expected, within reason. The dataset collected by your Garmin's Barometric Altimeter will be fundamentally different than map-based elevation that is corrected by an elevation database. 

    0
    评论操作 固定链接
  • I do not really expect a 1500 ft difference (5800 ft vs 7300 ft) between the two devices. The numbers were somewhat closer between the iPhone and the Garmin before the auto correction was applied. If the algorithm is based on measured, on the ground topology, when I apply the correction to the Garmin data shouldn't it be the same as the corrected Iphone data over the same route? After applying the elevation correction to my Garmin data, there was only a 0.5% difference when compared to the raw Garmin data, but still a 500 meter difference to the iPhone data. Hope this makes sense. Thanks, Gary
    0
    评论操作 固定链接
  • Doh! When I rechecked the numbers the corrected iPhone data was the same as the corrected Garmin data. Now I just have to figure out which is more accurate. The barometer or the mapping data.
    0
    评论操作 固定链接
  • Why does the strava route builder not deduce the same route elevation as the corrected version. Rapha rising challenge requires977m of vertical gain per day. I used the route planner and came up with a route that has 865m of elevation gain. I rode the route and my phone (Samsung s4) reported 878m, then the ride was corrected by strava to 765m. If strava is using topographical database to correct ride data (reasonable considering GPS elevation inaccuracy) why does the route builder not come up with the same total?

    0
    评论操作 固定链接
  • HA!!! Your lucky James. My elevation showed at 800 metres. That was after loading. I looked again 5 minutes later and it had been smoothed to 97 metres elevation gained.

    I loaded the same file to my Garmin account and it showed 637 metres gained.

    Now I know the segment I road and according to the Strava site the elevation gain on that segment is 27 metres and I was riding repeats from a little lower on the slope. BUT even if I was to accept that it was only 27 metres since I did 27 climbs that should still be a minimum of 675 metres.

    Looks like I'm pulling out of the Rapha Three ranges challenge.  At this rate I'd have to climb at least 53,000 vertical metres according to the Strava site to qualify. 

    0
    评论操作 固定链接
  • Hi Elle, thanks for the explanations. These make sense and I now understand the differences in different GPS devices' altitudes - thank you. However, I would always expect corrected GPS figures to match  the altitude figures in the Strava Route builder as surely Strava uses the same corrected data for these functions? This is the Strava app so we should all be using the same corrected figures and playing off a level field - Pun intended ;-) I am new to cycling, loving it and the challenges in Strava. I am using your route builder for a climbing challenge which is tough enough but so far all of my rides are getting "corrected"/reduced - sometimes by 25%. This is in comparison to your own route builder. Help!? 

    0
    评论操作 固定链接
  • I'm having similar elevation profile - related problems, that are driving me crazy. I'm recording with SGS3, which has barometric altimeter. Resulted GPX files represent the elevation profile pretty accurately, but then Strava corrects it with it's 30x30m elevation database and screws it baaadly. Often I get "downhills" (sometimes even longer than 100m) on the tracks, that were actually pretty steep uphills only.

    If I may suggest, PLEASE make an option to either show elevation profile from track or database corrected one. Maybe it would be even better if You could always show both (in different color) and also show elevation gained for both.

    I put a lot of effort in creating quality segments, but most of the time they have elevation profile not even close to reality because of this "corrections" ;(((

    0
    评论操作 固定链接
  • Hi Elle,

    I experience problems with inflated elevations calculated by STRAVA, here is an extreme one with an error >100%: http://www.strava.com/activities/266235362
    This hilly ride was logged with MotionXGPS on an Iphone.

    MotionXGPS calculates 
    Ascent 2.060m, Descent 2.049m which seems to be accurate.

    When I simply add all ascents and descents in the GPX file, Excel calculates
    Ascent 2.641m, Descent 2.634m, which seems ok also since its without smoothing.

    Strava however shows an elevation gain of 4.340m! 

     

    There seems to be a systematic mistake somewhere in STRAVA's server's algorithm. also this one http://www.strava.com/activities/266575582 is inflated. STRAVA=4.064m, MotionX = 2.940m.

    Interestingly the graph seems to be accurate, when you estimate the gains from the graph you end up around 2.900m.

    Cheers

    Charly

    0
    评论操作 固定链接
  • I've Samsung Galaxy S3, which has barometric sensor. When I record my rides with my phone using Strava Android application, I'd love if the app would also record barometric data from the phone, and use that for elevation instead of Strava's elevation database.

    0
    评论操作 固定链接

请先登录再写评论。

没有找到您要找的?

新建帖子