F5 - Barometer settings/modes/issues

So there it goes ;-)

A bit disappointed that there is :
- no altimeter calibration using local MSL pressure from an airport (calibration using known or GPS altitude only)
- no fine tuning of barometer so that it matches the local MSL pressure after the altimeter has been calibrated. In my case MSL pressure is 1020.6mb and the F5 is showing 1022.1...

Am I missing something ?

On the plus side the barometer module can be set to :
- Altimeter : it assumes constant weather so any changes in pressure will be altitude related
- Barometer : it assumes constant altitude so any changes in pressure will be weather related
- Auto : it selects the mode based on "movement" per the manual. Not sure if that's GPS or accelerometer movement ?

UPDATE : after a week no issues with erratic Altimeter/Barometer values like those reported on the F3HR

F5 compared to Motorola X :
  • You are making the assumption that ridewithgps is giving you the correct total ascent/descent numbers. Ridewithgps is using either GPS elevation which is notoriously ‘noisy’ (so adding spurious ascent/descent because of the random up/down noise) or it corrects the elevation to a map during post processing, which can also introduce errors and additional ascent/descent. An example of corrected elevation incorrectly adding numbers is when you cross a valley on a road bridge; the road bridge is flat but the corrected elevation follows the actual ground, so will incorrectly add the drop in to and rise out of the valley to your ascent/descent numbers. A barometer, on the other hand, is very good at tracking your ‘relative’ altitude changes but is susceptible to error in absolute altitude as the ambient pressure changes due to weather. Nothing is perfect, but I generally find barometers give a far more ‘precise’ elevation profile and reliable ascent/descent figures than the other methods, even if the absolute altitude has drifted between start and finish, thus has become ‘inaccurate’ (particularly noticeable if you start and finish at the same point).

    The 50 ft drift you describe whilst staying in the same location for hours is not unusual. The watch does try to ‘lock’ elevation if you are not moving, if you have the right settings (Not During Activity to Auto https://www8.garmin.com/manuals/webhelp/fenix5/EN-GB/GUID-CDC9DDEE-C576-4E1D-B007-BC737686793A.html ), so that changes in pressure are attributed to weather rather than altitude; but ultimately that ‘lock’ may not work if you’re moving your wrist sufficiently that it thinks your active and moving.

    Finally to Auto Calibrate. Unlike Suunto watches with barometers, Garmin haven’t implemented a continuous calibration on the Fenix 3 or 5 lines, where the watch continuously checks the drift in absolute altitude from barometric altitude against GPS and corrects once there’s a big enough difference that is outside the +/- accuracy envelope of GPS derived altitude. The Fenix 3 did work like this upon release, but the feature was changed in later software updates. So now the Fenix watches only Auto Calibrate once at the start of an activity in a rather complex way that Garmin have never written down and published. The best description from observations of others on these forums is given by PubBike here https://forums.garmin.com/forum/on-the-trail/wrist-worn/fenix-3/1249012-why-is-barometric-altitude-reported-in-activities-with-gps-enabled

    I thought I'd update how I think Auto Calibration of the barometric altimeter (BA) works based on more experience. Note that 8.30 is broken and does not work like this but 8.32 beta is fixed. If Auto Cal. is off, the BA value is unchanged at the start of an activity. With 8.30, the BA value is unchanged at the start of an activity regardless of the setting of Auto Cal.

    There are three distinct cases:

    Case 1. At a given location, if the BA was previously manually calibrated during an activity at this location, or is now or was previously calibrated using the Settings menu after green ring (GR) at this location
    - at GR, the BA will be set to the elevation of that calibration (thanks mcbadger for figuring out the Settings menu bit)
    NOTE: It is not possible to access this list of saved calibrations and it is not generally known how long this list is, but it persists after soft reset (long press LIGHT).

    Case 2. If the watch is at a "Saved location" and the BA was not previously calibrated at the current location during an activity or after GR using the Settings menu
    - at GR the BA is set to the elevation in the "Saved location".

    Case 3. If the watch is not at a "Saved location" and the BA was not previously calibrated at the current location during an activity or after GR using the Settings menu
    - at GR the BA will track the GPS elevation until START is pressed at which point the BA will be calibrated to the GPS elevation at the start of the activity.

    All cases: If the activity is started, the BA will be set to the new value. If the activity is cancelled, the BA will revert to the value prior to the green ring appearing unless the calibration was done in the Settings menu during the start of the current activity.


    PS: Maybe the reason why this isn't in the manual is because Garmin gave up trying to explain it in remotely simple terms because it is too complicated.
  • Well, o.k. Comparing against someone else's elevation calculation isn't good, but now I can compare the Fenix 5 to itself. I did one of the same rides from my post above again today, exact same ride. Here's what I got:

    https://ridewithgps.com/trips/19191247 - Old ride, +2185 / -2162
    https://ridewithgps.com/trips/19276656 - new ride, +2061 / -2003

    I thought I had auto calibration turned off on the altimeter and the barometer in altimeter mode but I just looked at the watch and this was reversed! I certainly didn't make that change, but then again I have checked this after syncing with the Garmin Express which might have pushed these changes. I'll have to check that.

    Still.... +124 / -159 is a pretty big difference between two identical rides. Any thoughts on that?
  • "a pretty big difference" ? What did you expect ? It's 5% for a three hour ride during which there are bound to be weather changes. It it perfect ? No but that's the closest it gets and is way more consistent that GPS elevation or "map based" elevation.

    By the way the barometer settings have no impact (as it should be) on D+ during an activity.