This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Elevation / Ascent errors on long rides/runs - elevation slow to react

*** appears to be fixed in Beta fw 9.73 ***

On 2 to 3 hour rides/runs, I have started to notice after about an hour, the elevation is getting heavily smoothed on the 935, like the BP gauge is being slow to react, causing the Ascent / Descent to be way off. You probably wouldn't notice it without another device that is recording correct values. Also, Just looking at your elevation plot on GC doesn't show enough detail for it to stand out. But if you display the elevation plot over each other with software that actually shows full detail, it is very obvious. In the below graph, I am comparing it to the Edge 130 but I have also seen this behavior compared to a 1030.

In the graph, the lighter green is the 935, the darker the Edge 130. both are set to every second recording. You notice how much smoother the 935 graph is especially as you get closer to the end of the ride. It doesn't capture any of the detailed ups and downs. The section highlighted in yellow is an out and back and makes this issue pretty clear. On the way up, you'll notice the 935 and 130 match almost exactly. On the way down, the 935 doesn't pick up the100' "bump".

I thought it might be due to sweat on my wrist clogging the BP ports so in this ride I had both mounted to handlebars (the 935 using the Bike Mount kit).

For this particular ride, the 935 reported an ascent 2986', where as the 130 had 3645' (which matched almost identical to corrected elevation).

FR935: https://connect.garmin.com/modern/activity/2809486628
E130: https://connect.garmin.com/modern/activity/2809486664

Has anybody else seen something like this? ciq.forums.garmin.com/.../1368507.gif
  • +1 I did meet the same issue with a previous software release 8.0. After several discussions with garmin and several reset :( Garmin suggested a replace of the 935.
    I made a 120km bike ride with my edge810 and the 935... 2135m of positive elevation with the Edge and only 935m with the 935. On Garmin Connect, if you active the correction, the positive elevation increase to 2000m... definitly the 935 goes wrong.

    Please Garmin... Can we consider an auto reset of the barometer when starting a ride and then a cyclic reset of the barometer for long ride? Isn't it a cheap corection?
    Most of us are really disapointed with 935bugs. Thanks
  • Just to beat a dead horse, I just went back through my workouts right around the upgrade to 9.1 (verifying in the FIT file the FW version). I upgraded from 8.0 to 9.1 on June 7. Every activity on 8.0 was perfect when compared to an Edge 1030, Edge 130, or Epix. Every workout on 9.1 and 9.6 has the issue. So it is pretty much 100% guaranteed 9.1 introduced the issue. This doesn't mean every device has the problem. But it does mean Garmin did something with 9.1 that caused it on effected devices.
  • I have similar case.
    Elevation gain is 667 m and elevation lost in 1201 m, but some elevations are very steep (over 14 % and speed below 6 or 7 kmh) !
    1 to 20 meters difference between lost and gain is not a problem for me, but so much?

    Is there something wrong with setting Smart or 1 s(ecund)?
  • Same problem here: today I paid specific attention to the tracked height during a bike ride. On uphill segments height sometimes did not react at all or even decreased (!!!). On downhill passages again, height sometimes increased first before it decreased later on. Very strange.

    In the next days I will compare that with a GPSMap 64s to check if this is a specific Forerunner 935 issue, a general Garmin issue or if barometric altimeters simply aren't reliable unless you spend more money on precision sensors.
  • I have a new theory to this. Looking back at my logs, it is my rides that this problem really comes into play. Looking at my runs (I only have a couple where I also had my Epix for comparison), I don't see much discrepancy, even with mountainous terrain. I wonder if there is a software "feature" where if they don't detect arm movement, they think you are stopped, and thus switch to barometer mode assuming changes in pressure are due to weather rather than elevation. And in certain biking conditions, you don't have much arm movement. My rides have all been on the mountain bike so this is a little questionable.

    Can anyone confirm they have been seeing this on runs as well? Or just while biking? As a note, I have not found my Edge devices to be particularly accurate while I run.
  • Every run is like this or even more ups and downs are smoothed out. First a normal 10k and below that a hill repeat.
    I think its a smoothing algorithm that is not working as expected.
  • Every run is like this or even more ups and downs are smoothed out. First a normal 10k and below that a hill repeat.
    I think its a smoothing algorithm that is not working as expected.


    I assume the 935 is the red in the first one and blue in the second? What is the comparison device?

    And how do you have your 935 activity configured? Is it configured as a "running" profile? I somewhat wonder if the type of activity profile could have an impact. I don't know what special computations Garmin does based on activity. I am usually using a CIQ app to record my activities but specify to the FIT file that it is a cycling activity.

    It seems like more people would be seeing this if it happens just under normal use so I'm wondering what we all are doing that's different. I have my new 935 so will wear both on tomorrows run with 1440 feet of climbing along with my Epix. Then I'll do a mountain bike ride on Friday where I have previously seen a big discrepancy.

  • Sorry, forgot that info.
    its a compare between Polar V800 and a fenix 5X and you assume right about the colors. The 5x is new after I got a new one due to the elevation/barometer problems. It didnt help of course, but did run the v6.83 for a while and then the 5x elevation was better than v800.
    I always run with both devices and merge the data that is best.
    After all discussions on this, I think the issues been the same in 935 and 5x.

    Ive been using the run app.

    I think you see bigger diff in total ascent elevation when you move faster, like biking, due to that its more often that the elevation dont recap the change.
    If you hiking its doing a better job. This is jog/hike from 2 weeks ago. Purple V800 and green 5x.

  • Former Member
    0 Former Member over 7 years ago
    I have the same problem. The elevation line seems to be smoothed. After an hour of biking small ascents like +5-10m / -5-10m are not recognized. On a 50 mile ride with 1100 hm over the whole distance the 935 counted 624m up and 596m down. The Edge1000 counted 1110m up and 1106m down, which is almost the correct value of this course. This problem is on all rides longer than 1 hour. If i look my activitys it seems to happen since software version 8.0. On mountaineering its the same but the elevation difference to the correct value is lower because its often only one big ascent. Please Garmin fix that!!
  • I did a ride last weekend with my Egde 520 and 935. I was keeping an eye on the elevation gain throughout the activity. For the first few hours they were within a few metres of each other, then the 935 started lagging. Like weizn, I noticed that climbs of 10m or so were not registered on the 935. By the end of the ride the 935 had lost 500m over the 520 (2400m vs 1900m). As a cross-reference, my wife's 520 also reported 2400m.

    It appears as if Garmin are using some sort of dynamic filter coefficient on the 935's altitude. Not sure what the conditions/situation is for it to go into this over-smoothed mode though.

    We're not alone - lots of complaints about it on the Fenix forums too.