Comments

42 comments
  • So this old bug now also applies to the iPhone 6S and 6S+, as Apple unsurprisingly continue to include barometers in their phones. I just bought an iPhone 6S and was pretty disappointed to find Strava is throwing away the barometer data, and continues to massively overestimate my elevation gains over bridges.

    I'm wondering if anyone's figured out a workaround using another app to record the activity in a way that encourages Strava to actually use the elevation measured by the phone?

     

    Also, Strava, you might want to update your FAQ if you're intentionally not using the barometer, it's a bit misleading for those like me who took it at face value... otherwise this really doesn't seem like some complicated or niche feature that should take over a year to fit into the development cycle.

    https://strava.zendesk.com/entries/20965883-Elevation-for-Your-Activity#whyrecalc

    • Device with barometric altimeter:  If the GPS device used to record your activity has a barometric altimeter, Strava will use the elevation data in the original file. We do some smoothing to the data, which includes discarding outliers to reduce noise. The amount of smoothing is less on activities with barometric data than it is for activities without barometric data. If your device has a barometric altimeter, but your elevation is being recalculated, please refer to this FAQ.

     

  • We don't currently use the barometric sensor in the new iPhones because it was not accurate enough. Please let me know if you have any other questions

    "They must be kidding. I don't think any sensor can be less accurate than currently used method. "

     

    What they probably meant to say was "In the limited testing we did in unmountianous regions where we happen to be, the iPhone's barometric altimeter was not as accurate as our elevation database, which we fill in from after throwing away the track's elevation data. The barometric elevation data is indeed more accurate than the GPS elevation, but not better (in our small and wholly insufficient testing) than our elevation database, so not good enough for us to not override it the way we don't override other barometric units."

    That's my guess, at least. Where I ride (the mountains of Kyoto, Japan), their elevation database is laughably incorrect, so the iPhone barometric altimeter would be a welcome step up. IIRC, the "Runmeter" (AKA "Cyclemeter") app lets you use the iPhone's barometric altimeter, though I don't know whether Strava then respects it or overrides it.

  • I have receiveid reply from Strava (to my personal ticket) regarding usage of iPhone barometric sensor data:

    We don't currently use the barometric sensor in the new iPhones because it was not accurate enough. Please let me know if you have any other questions

    They must be kidding. I don't think any sensor can be less accurate than currently used method. I get ~200m elevation difference compared to Garmin on my 50km ride. Pretty sad.

  • It would be useful a list of all the devices with a Strava supported Barometer (ok Garmin, but also iPhone6, Nexus4, Nexus5, ...)

  • Would someone (Strava dev/support) be able to post an ETA on the iPhone 6 Barometer support? I (and I'm sure many others) have no real reason to drop 300+ on this device unless this support is added and confirmed.

    Thanks.

  • I just did a test this morning, it looks like Strava (and any other app) doesn't need to do anything special to get the more accurate elevation data on the iPhone6.

     

    I rode this morning with another person who had an iPhone5C and I have an iPhone6, see the compared elevation profiles that are attached.

     

    iPhone6 Activity: http://www.strava.com/activities/197070628

    iPhone5C Activity: http://www.strava.com/activities/197074763

  • Anyone from Strava care to weigh in?

  • @Andrew - we should be normalizing elevation for the Golden Gate Bridge (as well as other bridges), but there's a bug in the way that we're working to fix. Hopefully we'll have more info soon. 

    @Bruce - we're aware that especially for elevation data outside of the US there are some difficulties with resolution for our elevation database. We use a publicly available database of surveyed elevation which is the best that's available right now. Only trouble is the resolution is 30 meters. Think of dividing all land surfaces into 30 meter squares and determining the average elevation value per square. When your GPS coordinate falls into that square, the GPS point is assigned that average elevation value. Particularly for rugged terrain, this can produce a lot of noise in the data. Our apologies that we don't have a better service to provide at this time, but we have a larger project in the works to address this. 

  • Thanks Elle!

    I really appreciate your continuing to follow up on this. There are probably a hundred post/comments in the Strava forum and elsewhere on this issue..far more than any of us have time to read. I think you are first at Strava to be aware of the extent of this issue.   In just the Million Foot Club alone, here is just one thread (which gets heated) on this exact topic and more links I included at the bottom:

    http://www.strava.com/discussions/club/47976/posts/164156/comments and here are highlights:

    Million Foot Club Discussion Topic

    • George Vargas REV Cycling
      George Vargas REV Cycling 1/13/15, 7:59 PM

      At the beginning of the year I made a comment about the overinflated elevation gain when using iPhone Strava App as your computer.  I offer you a classic example of overstated elevation gain.  https://www.strava.com/athletes/702924 I makes it even worse is this athlete truly believes that he is climbing for example over 16,000 feet in 55 miles https://www.strava.com/activities/238585424

    • Phil Millham
      Phil Millham 1/13/15, 8:35 PM

      Maybe Erik actually doesn't understand what 'elevation GAIN' is!! Erik, you can only count the 'gain', NOT the descent as well. :) Okay, all you peeps using iPhone or Android apps, keep using them. Keep fooling yourselves. Just promise me you won't use them in a Strava climbing challenge. Because to any sane thinking person, that is known as 'cheating'!!

    • Eric Foltz
      Eric Foltz 1/13/15, 8:54 PM

      On the second ride he posted you can just use the elevation profile to see that he only climbed around 7000'. The only accurate way to measure elevation gain is by using a GPS with a barometric altemeter that self calibrates while you are riding. The phone app usses map data then runs it through a smoothing algorythm. It usually end up with a 25-50% innaccuracy which can be compounded if the person then uses the "correcting" feature after the ride is uploaded.

    • Mike 'Mr. Crash' Solis
      Mike 'Mr. Crash' Solis 1/13/15, 10:18 PM

      They're so grumpy!

    • Douglas Kubler
      Douglas Kubler 1/13/15, 10:57 PM

      Users of phones and non-barometric devices do not have a correction option on their page because it has already been done for them.  The map database lookup is the correction process.  When a users selects "correction" for his barometric GPS he is saying "throw away all the elevation data and pretend I'm using a phone".   You will see some difference between a phone and a corrected GPS but that's because the phone has less horizontal accuracy. 
      I'm sure some naive people believe that using a phone or correction is an honest choice.  It's the rules of Strava and Strava refuses to change the rules.

    • Douglas Kubler
    • D. Kreature
      D. Kreature 1/14/15, 7:28 AM

      His data says San Elijo Road is a 42% grade in places. For comparison, here's a picture of what's considered one of the world's steepest streets (35%): 
      http://en.wikipedia.org/wiki/Baldwin_Street#mediaviewer/File:DunedinBaldwinStreet_Parked_Car.jpg

    • Douglas Kubler
      Douglas Kubler 1/14/15, 7:42 AM

      The data you see is what Strava derived from an ordinary ride.  For another example look on Facebook. (link above)

    • D. Kreature
      D. Kreature 1/14/15, 8:36 AM

      Right, Doug. I haven't ridden it, but from what I gather San Elijo Road is maybe 17% at its steepest. To George's point, the ride data seems highly distorted.

    • Brian Toone
      Brian Toone 1/15/15, 6:17 AM

      I've been thinking a lot about this since the last climbing competition. One of the things that bothered me was that there would be 25% downhill in the middle of a climb and no increase in speed. Similarly, there would be a 25+% uphill in the middle of a 5-6% climb and no decrease in speed. You can see this as huge power spikes in strava's estimated power for a ride. The speed data could be combined with the elevation data to add a better smoothing filter to elevation corrected rides. Obviously, there are some exceptions where you are braking hard on a steep descent for example: http://youtu.be/DbgtT8N-nGU but for "micro" descents in the middle of a climb, if your speed doesn't take any bump up at all then throw out the descent completely (i.e., keep the elevation the same) until the next increase in elevation. It appears Strava is going to have a monthly climbing challenge this year so if nothing is done about it. There will probably be more wide spread complaints from people who don't normally participate in the special one-off climbing challenges.

    • Douglas Kubler
      Douglas Kubler 1/15/15, 9:57 AM

      Strava's correction is worse than doing nothing.  The original data from the above mentioned 16000 footer could be a more realistic 8000 from pure GPS. Add a smoothing filter for randomness and you're done.  Your monthly competition, Fabio, has a 9.3 mile descent (segment). The track from an Edge 500 shows a constant descent except for one small bump - as a mountain road should be.  When Fabio (and others -phones or corrected gps) ride it they get credit for climbing 2294'. The climb back gets a boost too.  Find me on Facebook for another case with pictures.

    • D. Kreature
      D. Kreature 1/15/15, 2:10 PM

      The bottom line is if you want credibility in a climbing club, or a climbing competition, use uncorrected barometric data from a Garmin. I understand why riders are reluctant to give up the iPhone Strava app: it's free, it's easy to use, and it's a single device. Garmin's only advantage is its accuracy.

      One hurdle to Garmin adoption is expense, so I contacted Garmin, said we were a climbing club that wanted all our members to get legit, and asked if we could get a group discount. They basically shot me down, and referred me to Sport Chalet. I'll let you know if Sport Chalet comes through.

    • George Vargas REV Cycling
      George Vargas REV Cycling 1/15/15, 6:42 PM

      D. Kreature I own a bike shop I can help you and this Million Club with pricing.  The margins are slim but if there is volume then there is an opportunity.  my business email is revcycling@gmail.com for those that wish to contact me for a Garmin purchase.

    • Sam Stevens
      Sam Stevens 1/16/15, 3:28 AM

      I use an 800 and think it is pretty accurate. One issue I am sure most are aware of is when riding in rain or low pressure days, readings are off and way low. That is when I once in a while use the auto correct. In my opinion the Strava auto correct seems fairly close.

    • Douglas Kubler
      Douglas Kubler 1/16/15, 7:31 AM

      Sam, For climbing and descending the absolure pressure (altitude) is not important.  The important data is the change in altitude. If you want your 800 to have accurate absolute readings enter locations with known (surveyed) altitude into your 800. Your usual ride start point is a good one.  See page 26 of your 800 manual "Editing Locations".

    • George Vargas REV Cycling
      George Vargas REV Cycling 1/16/15, 8:46 AM

      Hello Sam Stevens yes during poor weather the 810 also had inaccurate data throughout the ride.  I used auto correct on one such double century with rain through most of the 13 hours and got fairly close about 1K off Should have been close to 15K was only 14k

    • Douglas Kubler
      Douglas Kubler 1/16/15, 9:39 AM

      You can't always go by advertised gain.  I had a conversation with a promoter about his data.  He said that people want to sign up for events with big gain.

    • Ian Douglas
      Ian Douglas 1/19/15, 3:54 PM

      Strava now has an extensive database of ride data accumulated from all its users' many rides over the same segments, so why don't they use their own collected data to smooth out their app's data? They must have ten of thousands of instances of rides up various climbs with all kinds of different devices, so they should have their own ride elevation database by now.

      Elle...there are many more.  This was just one thread.  Here are some more:

      http://weightweenies.starbike.com/forum/viewtopic.php?f=14&t=122058

      http://forums.roadbikereview.com/general-cycling-discussion/garmin-strava-elevation-discrepencies-285893.html

      http://singletrackworld.com/forum/topic/strava-altitude-gain-issuesv-garmin

      https://forums.garmin.com/showthread.php?37042-Elevation-Differences-between-GC-and-Strava-with-2-50   and the list goes on!

      Thanks again!

         Bruce

  • Elle, it's time for the mobile team to move on this. The API has been available for nearly a year, with WWDC looming. There are those among your users who have used the API and may have trouble understanding the impressively long lead time on having more accurate elevation data on what is, in many ways, a class-leading app.

  • Hello, Strava, any updates on this topic??

  • From what I have read, the barometer is quite accurate.
    For example:
    http://www.extremestorms.com/iphone_6_barometer.htm

  • Using the barometric altimeter on the iPhone would make Strava my favourite sports app + gadget replacement! It's just frustrating to see the full potential not being tapped. This would be a win-win situation: Broader refinement of the "elevation look-up serveice" as well as quality elevation data in activities for users of "recent" iPhones.

    Hope it will be implemented soon!

     

  • Interesting. It seems that the barometer has some effects (the smoother profiles of the Iphone6 seems to me more realistic).

  • Correct, the iPhone 6 elevation data reflects the actual elevation much more accurately.
  • Although climb data may be better in the feed, it would be even better to be able to see the altitude / climb stats in the Record screen of the app while riding.

  • At the moment there is no use of the barometer. 

    "Hello,

    I apologize for the inconvenience. Currently this feature is unavailable however stay tuned for future updates!

    Cheers,
    Rudolph
    Strava Support Team"

     

  • Since the last update was Oct 14th, I assumed Strava added Barometric support for the iPhone 6 M8 chip so I emailed Strava support to confirm and unfortunately, here is the response:

    • Hi Bruce,

      We are still not accepting barometric data from the mobile app at this time. Instead, the GPS data gets cross-referenced with an elevation database.

      Sorry about that!

      Best,
      Ashley
      Strava Support Team

    • December 01, 2014 03:49 PM

    A competitor of Strava added Barometric support on October 25th 2014...why is Strava so slow on this enhancement?  Total elevation is typically off by 50% or more for me and many others.

  • That sounds great. I'd love to be able to use my iPhone for that info. Which competitor?

  • Cyclemeter (http://www.cyclemeter.com/) claims to have support, I haven't personally used it yet, but many people have reported it as extremely good.  Cyclemeter will even upload the data to Strava for you automatically after the ride.

  • I just got the app and will try it out this afternoon. You need the 'Pro' in app purchase ($6.49) to turn on the barometric altimeter
  • I just heard back again from Strava support that barometric support is apparently not even on the roadmap.  Very surprising.  When I watched the Apple iPhone 6 launch back in Sept '14 I remember thinking Strava would immediately provide barometric support (M8 chip) due to the horrible elevation data they provide.   Here is the text from Strava support: " if there is enough interest, it may be something we put on the roadmap."  The solution for now is as Justin describes above.  Buy CycleMeter Pro which can automatically import the data to Strava.

  • Even though Cyclemeter records an excellent barometric elevation trace, at present Strava won't accept it anyway:

    "At this point in time, Strava will recalculate the elevation for any files that come from devices not on our list of 'accepted barometric devices'. This includes mobile uploads.

    We know this isn't ideal and are looking into ways of improving this data in the future.

    Sorry about that!

    Best,
    Ashley
    Strava Support Team"
  • Rudolph from the Strava Support Team,

        It's been 4 months since you mentioned in this forum (below) "stay tuned for future updates!" regarding support for the barometric features support in the iPhone.   I'm hearing that support is still not planned.  Why is this not being addressed?

  • Any update from Strava support on this?  Every Strava update gives me hope that this 50%-100% (too much elevation) will be fixed and I'm continually surprised Strava doesn't address it.  Thanks in advance for your feedback on the status.  It's now 6 months since Apple provided barometric support but Strava has not used it.

  • Just checking in to see when we can finally expect this feature to be added so I know whether to continue my Premium membership or not.

  • The elevation data in Strava has so many errors (Golden Gate bridge is not 0 feet above sea level, thank fully).  Please use the barometer data from the mobile uploads, or allow us to override it. Please. :)

  • Interesting comparison for the 2 iPhone models, obviously the 6 has more accurate GPS data, filling more points thus being able to reference more altitude points along the route. I'll see if any of our rides with various phones compare with Garmin data.

  • Thank you Elle for the feedback.  The Strava developers must not be aware of how bad the elevation errors can be.  50%-100% is the norm for me and you can tell by the sawtooth profile when all the climbs are smooth.  They should join the Million Feet club so they can be more aware of how bad the errors are.  There are many angry rants about Strava and also sometimes blame iPhone even though Apple fixed the issue by adding barometric as of iPhone 6 but Strava has not implemented it.  I can't think of any other products that allow their accuracy to be so poor.  Can you please ask them join the club and read about it?

  • Thank you again Elle.  I thoroughly appreciate the interactivity!  The comment "We use a publicly available database of surveyed elevation which is the best that's available right now" is good to hear since you have a much simpler solution.  You no longer need to use a database since the source comes directly from the barometric reading from the iPhone...you use the new API call for barometric.  I tested with CycleMeter (which added iPhone Barometric support back in October 2014) and the elevation profile is perfectly smooth just like the 6% grade of the road.  So its not a matter of finding a better database of elevation data.  Strava Premium customers have been very patiently awaiting this fix for almost a year now (since iOS Beta) and still nothing.  Please be customer focused.   My sincere thanks again for the collaborative discussion.

Please sign in to leave a comment.

Didn't find what you were looking for?

New post