Altimeter performance test

The test setup was as follows: To compare the FR955 (worn 24/7) with two other watches:  around 19:00 I start wearing a Fenix 6X Pro and a Fenix 5 Plus on my other arm. Around 20:00 I go for a run, about 1 hour. Then I rinse all three under the tap, towel-dry them and put them on a shelf until I'm done showering, then put them all on again. At 23:00 I look at the altimeter and barometer graphs, make screenshots and take off the two Fenixes. 

The first test showed markedly different behavior of the FR955, but after a second test all three showed nearly the same erratic behavior, wrongly attributing ambient pressure changes to altitude changes.I don't think all three devices are broken in the same way.

I think the Auto setting attributes all ambient pressure changes larger than approx 1 hPa per hour, to altitude changes. And this behavior is common on more devices, although some probably have slighty different thresholds than others.

Changing from Auto to Barometer Only solves this issue. I am happy to trade in altimeter functionality for a good barometer. 

Thanks for the description of your setup.

But this seems fine. So in the next tests you could not reproduce the strange behaviour of the forerunner 955.

I also encountered a drift in my activities - as I have written. And it is also pretty hard to filter what is the change in the ambient pressure and what is due to altitude change. Maybe they have some standard model of the 1hPa per hour - or nothing. But they could also work with weather models as long as the device is connected to the phone or to wifi at some point (what they probably don't do...)

So for me the auto setting is working fine so far, but as you say it depends on the usecase. Nice would be if the altitude determined by the GNSS systems would also be logged in the files. Then you have at least a backup if you are interested.

/resized-image/__size/640x480/__key/communityserver-discussions-components-files/648/5417.image.JPG

uploaded picture as a link since forum was giving errors when I tried to upload picture

How long have you had your watch? Your altimeter seems pretty accurate. I have been having a different experience and my start and end points are never the same elevation. If I'm lucky they are within 30ft, but this run had a 150ft difference over 10mi. The elevation already drifted up 60ft by mile 1.75 since I did a short out and back at the beginning and passed the trailhead again.

The red line is the original elevation from the device and the green and blue lines are the corrected elevations from Garmin and Strava. The green line for Garmin corrections has more noise and don't accurately reflect the actual elevation profile (especially at 3.5mi and 8.5mi). The blue line for Strava corrections appears to follow a more similar profile to the original device elevation profile but shows a 400ft difference in elevation gain (1815ft vs 2244ft). So in addition to not showing the correction elevation, it also likely overstates the elevation gain.

I always calibrate it using the DEM calibration before starting a run (even though that shouldn't impact how much it drifts).

I'm wondering which units you use for altitude. It seems that your language is French. But your numbers are very strange ...

units are meters. my altimeter is completly *** up. today, i try not to use barometer for ski mountaining. it is better. only 27000m uphill instead of 1900. i disabled 3D distance and speed, so, distance is ok now.

I got is for Christmas two years ago - so it is 2 years in service now and I'm still happy with it.

I fully understand that it is annoying, when the altimeter is not working propery.

Did you use it often in the water? Because there are hints also in the forum that pool water might degrade the barometric sensor pretty fast. I'm not in the water very often. Think mine saw pool water around 20 times for about one or 2 hours and sea water 10 times for some minutes and a shower each day. I only had it off for charging in the last two years but my "water time" is still low.

Also there might be devices which are more temperature sensitive. Because I did not encounter a huge drift correlating to the device temperatures than others.

Concerning the corrections, in my opinion it is still difficult here to get the proper DEM data. So when climbing a mountain it is fine because it knows the altitude in the valley and at the summit. But when running on hilly terrain the quality of the map correction is not very good.

Anymore tests about this?! My first 955 Solar was replaced because Garmin think that the sensor was broken. They sent me another one, but in the past 4 months the symptoms of wrong altimeter has begun again.

It's so annoying that a what that cost me around 700€ don't give acurrate data.

I am still having the same issues. I am still noticing most of the altimeter jumps happening when I go from warmer temperatures to colder temperatures or vice versa. Which leads me to believe that Data is right that the algorithm or calculation is off when the watch is trying to account for the temperature changes relative to altitude (at least that is my understanding of what Data said...).

I can not confirm the temperature depency in my case. Last weekend I did crosscountry skiing, starting directy out of the car and getting out to -8°C ambient. Results were in range for my case.

1 hour time, about 300m vertical gain / loos / 14km distance and a difference at the round trip 3m.

hpaixao Do you often go swimming with the watch?