Wildly incorrect spiky altitude data

Mff73

@sesam
Ok, we missed this info that watch also recorded 600m ascent (which was not the case, at least for me and the cloud issue).
but, but, but, … who knows with bugs
see what you can find and/or share the activity link if no privacy concerns.

sesam

@isazi I was thinking about this… I was wearing the watch the same way as always, over the sleeve of my jacket, using a 3rd party HR band. The only difference was that it was snowing like crazy, my clothing and the watch was covered in snow after a short while.

isazi

@sesam bingo! It happened to me under the snow a week or so ago, and something similar to @stromdiddily and @freeheeler I believe not long ago.

stromdiddily

@isazi I wonder if there were changes somewhere. I don’t remember this popping up as much last year…

isazi

@stromdiddily the release notes of version 2.13.18 says:

Fix some pressure sensor issues

But I actually do not know if it is related or not.

freeheeler

@stromdiddily
weather changes… we haven’t had that much snow in ages!!

It’s a bit of a drawback that we only have one sensor for measuring. Two sensors or at least channels from different sides would really really help here.

I know that Suunto uses Creo to design watches… maybe I should apply as a designer and give them a hand with the future baro watches

stromdiddily

@freeheeler well, it’s been raining pretty steady where I’m at since…forever lol

I really liked the location of the sensor on the A3P. I am not a fan of the side sensors for this very reason.

cosme.costa

@freeheeler said in Wildly incorrect spiky altitude data:

I know that Suunto uses Creo to design watches… maybe I should apply as a designer and give them a hand with the future baro watches

Don’t go so fast, if there is an opening I’ll apply first…Creo is my friend too.

freeheeler

@cosmecosta
maybe they need 2 great design engineers

stromdiddily

And then it totally redeems itself. Off by 5 feet at the top of the summit on today’s run.

lexterm77

@stromdiddily

I think it is better to just feed all data into tensor flow and give it as much as possible correct keys to learn how to deal with these conditions. I had better experience with S9 when biking than with A3P. Under certain conditions with S9B I have 600m ascent instead of 6m but with A3P I had predictable errors due to wind.