I was looking at Garmin's product support page for the 935. While reading about elevation, the website explained the "auto", "altimeter" and "barometer" mode settings in quite detail. In the section about calibrating the altimeter it reads "From here you can choose Calibrate to calibrate the sensor using a current known elevation or by using the GPS elevation reading, or toggle the Auto Calibrate setting on or off." Calibrating it manually with a known elevation or using GPS are self explanatory, but I can't find any info on auto calibration. Would anyone be able to explain what the auto calibration "on" does and how it does it? Secondly, why is there an option to turn it off? Is there a situation where one would want it on as opposed to off or vise versa? Any help with this would be appreciated.
For me is very annoying this problem. Here are two examples of short vertical activities. The correct elevations on top are: 2164m and 3167m. The graphs looks that because on the top i have made a manual calibration. Of course the gps elevation data field was showing the correct elevation, but the recorded activity is based on the barometric elevation, so the final D+/- is not ok :( Imagine how the final activity will look at a 10 hours activity with 3-4000 meters of elevation.
A few things : - Manual calibration during an activity has no impact on D+/D- - That graph actually looks very clean, there is no way you could get that with GPS altitude, it would have all kind of up and down "noise". - Yes it's not perfect but there is no better option out there right now than using the baro for D+/D-
You'd have to do a side by side but I doubt the FusedAlti graph and D+/D- would be any better. Again it's a software trick to show "corrected" altitude as if you calibrated manually every X minutes, it's not magic, they have nothing more to work with than the GPS altitude and the barometer.
Again it's a software trick to show "corrected" altitude as if you calibrated manually every X minutes, it's not magic, they have nothing more to work with than the GPS altitude and the barometer.
GPS altitude and ambient air pressure measurement is plenty to work with! I have already in another thread suggested an algorithm based on those two inputs which would do a lot better than just auto calibrating to GPS altitude every X minutes:
Continuously calculate a "shadow" barometric pressure from GPS altitude and measured ambient pressure.
Calculate a running average over x minutes of the shadow barometric pressure, perhaps combined with some trend prediction.
Use this running average for a continuous recalibration of the altimeter.
This would give an altitude which would be a lot more accurate than whatever can be measured with the GPS or the barometric altimeter in isolation. And the beauty of tracking the barometric pressure instead of tracking an altitude error is that altimeter recalibrations will not affect the tracked value.
I wonder if I should learn how this Connect-IQ stuff works so I could make my own data field, based on the above algorithm.
The pressure should not change with temperature changes in an open system.
So if you are in a house which has some vent openings to the outside, and you heat the house, the ambient pressure inside the house will remain the same. If you go outside the house, where it is cold, the ambient pressure will remain the same.
If you heat your watch, the heating of the watch will not change the ambient pressure around the watch - which is what the watch is supposed to measure. So if the watch changes its ambient pressure reading when you heat it, it is an error. But some measurement error probably can't be avoided. I have seen pressure variations of 0.3-0.6 millibar (equal to roughly 3-6 meter of altitude) when changing the temperature of the watch (a Fenix 3).
(There can be exceptions to the above, for example if you are in a very tall building and the opening to the atmosphere is at another elevation than you are. Then the air density difference on the inside and the outside can give a pressure difference if you open a door and go out. Also if you are in a building ventilated by force, there can be a difference between the pressure inside and outside. But in both cases you will also feel this as a push or pull in the door when opening it, and as long as you can easily open a door by hand, the pressure difference is probably only a few tenths of a millibar.)
Heating the watch probably affects the operation of the sensor itself rather than the air around it.
Yes, that is very likely (and I already mentioned that in another thread). But any larger error caused by temperature would render the measurements useless unless compensated in software.
So I am pretty confident that the pressure sensor is either intended to be adequately insensible to temperature changes, or its output is corrected in software before being used for anything. In either case, you should not see any significant altimeter error from a temperature change.
I explained above why you'd want to display both at the same time, that shouldn't be too difficult to "GET" either surely ? GPS altitude IS correct, barring any catastrophic environments, the only problem is that it changes a lot so messes up D+/D-, hence the inclusion of barometer sensors on GPS watches, and of course they're useful to anticipate/understand trends in the weather.
Uh, yeah, the same GPS that has problems once we move from wide open spaces to the areas in which the vast majority of us are using the watches? The same GPS that, while doing an admirable job, still goes off the rails in the horizontal plane? Got'cha. If you wrote, "GPS altitude is great out in the open, on top of a mountain, or in an airplane." I may have been with you.
