Fenix 6 Series - 20.42 Public Beta

5135968 said:

My opinion - and that's just my opinion, I have no insight here - is that this new altitude algorithm was actually crafted with the Epix / Fenix 7 and rolled into the Fenix 6 so they can maintain a common codebase as it's very late in the Fenix 6 product cycle to introduce this change otherwise. I doubt the programmers had much say in the matter.

sounds plausible, but I'm not sure why they couldn't take something what worked earlier from Fenix 6 instead, that would be more logical to me.

8250747 said:

After reading all the posts in this thread I still don't understand if it's worth upgrading from 20.30 to 20.42? With the altimeter and barometer everything continues to work exactly as it did at 20.30 or there are improvements. Or has it gotten worse?

Unless you want to experience bugs on yourself and report them to beta team, then of course not, because beta versions are for testers, not for users. At least in my eyes 20.30 seems to work better, there were no flat elevation lines, it has all the same features as 20.42. I see zero point in upgrading unless you have some troubles you want to resolve. Even if it seemingly got better, I'd still recommend not to upgrade to it before testing is over.

iNDEX3 said:

My apartment is at an altitude of 230 meters. But near the house the altitude is 217 meters.
I go outside and in the widget I see the correct height - 217 meters. I start the activity and see a request to confirm the altitude. But in the request, the altitude is not 217, but 230 meters. If I press "back" and start the activity again, the second time the height request is correct.

huh, that's something interesting and definitely worth testing. Thanks for telling! Do you use your watch in Auto mode and have calibration set to Auto or Nightly?

Myself, I have everything set on Auto and see fluctuations +-3 meters a day for my home, but it's not something what bothers me.

iNDEX3 said:

If I press "back" and start the activity again, the second time the height request is correct.

I see the same from time to time, but not consistently.

Ok started running at 42mt, 1hr with variable heights +/-50mts, back home at 44mt (I'm at 3rd floor so I should be at say 56mt usually). 

I used to see similar behaviour several firmware releases ago, some time last year.

I would leave home with something like 43-47m on the watch face (44m would be correct) drive 6km to my daily dog walk, when the watch face would track my descent down to between 10-16m (16m would be correct) and as soon as I opened the walk activity the altitude would leap back to 44-47m, basically exactly as it had been recorded overnight and not where the watch now thought it was.

It used to drive me nuts. Just such stupid behaviour. I don't understand how coding can be so bad.

Were you using Auto calibration or nightly?

It's a long time ago, but I think it would have been nightly calibration in order to get the calibration prompt (with the +30m jump) at the start of the activity. It happened so often that it probably forced me to switch to auto calibration overnight just to avoid the irritation and the need to fix it manually each day.

EDIT :  actually filed a report about this, almost word for word, in email conversion with Outdoorbeta on 8th December.

sensovision said:

sounds plausible, but I'm not sure why they couldn't take something what worked earlier from Fenix 6 instead, that would be more logical to me.

I think this altitude change is due to upcoming wrist based running power.

If you want decent wrist based running power (like Coros does now which basically makes a Stryd obsolete as Coros running power and Stryd running power are basically identical), you need to be able to measure elevation loss and gain accurately or at least differently to how it is now.

This is a guess but I'd be very surprised if this isn't a major reason behind the change in elevation algorithm.

eezytiger said:

It's a long time ago, but I think it would have been nightly calibration in order to get the calibration prompt (with the +30m jump) at the start of the activity. It happened so often that it probably forced me to switch to auto calibration overnight just to avoid the irritation and the need to fix it manually each day.

And yes, I switched from overnight back to auto as it was super irritating even for testing purposes.

5135968 said:

If you want decent wrist based running power (like Coros does now which basically makes a Stryd obsolete as Coros running power and Stryd running power are basically identical), you need to be able to measure elevation loss and gain accurately or at least differently to how it is now.

That's an interesting idea and that's the only reason I can see why they may backport the algorithm from Fenix 7 back to Fenix 6, otherwise it didn't make sense to me. Still, I doubt that F6 users will get any benefits in form of running power.

But yes, built-in running power is long overdue by Garmin so maybe soon we can confirm that your theory was right