Auto Calibrate Elevation to Location? Baro Altimeter?

How do you know that immediate elevation is from GPS rather than from the baro sensor?

If you turn GPS off the manually entered value remains unchanged.

After made all the possibles tests, i'm sure this latest firmware do not solve the problem. The altitute works and it's only correct when we start after made the manualy calibration.
Without that, the altitute it's 40/50 below from the real one.

Temperature an huge effect on altitude readings?

Most of my rides starts from home at 25m above sea level, and ends either at home or at my work place that is also at 25m. I always auto calibrate at the start of my ride. The displayed altitude is then correct. However during a 7 km ride the displayed altitude drifts, and when I arrive at work the displayed altitude is about 3m. When I look at the .fit file it has recorded both start and end altitudes of 3m. The recording is consistant, but 22m below reality.

However, yesterday I went out for a longer ride. I split the trip in to two halves, and corrected the altitude halfway. First leg, the altitude was about 12 meters below reality, but the second leg when I arrived home the altitude was correct. When looking at my .fit files the one variable different between these scenarios is the temperature recorded at the start of the rides.
- When commuting to work I wait a minute or two, calibrate the unit, and press start. When testing this today the start temperature was 18 degrees celsius, and at the end a more correct 4 degrees celsius.
- The second leg of yesterday's ride was started with the correct temperature as I started it after being outside for an hour. It showed my home at 27m in the end (within a margin of 2m). However the recorded file shows an start altitude of 2m above what was manually set, and the end altitude displayed in Garmin Connect and on the unit live at the end of the ride was the same. This indicates an +2m override that is to be seen in the recorded in the .fit-file, but is slowly integrated to the live feed from the Edge 820 display.

This points me towards two different variables for altitude error:
- Causing drift: Start temperature usually is way to high going from inside to outside, especially in winter. Have to wait 10 minutes for the temperature reading to be corrected.
- Causing faulty readings: The unit overrides the manually calibrated altitude.

Does this sound like plausible explanations?

Tried the same tests again. Rode the same round trip twice. The first time I started the Garmin while the temperature displayed was still decreasing after being indoors. The second time I started the tracking right after finishing the first, as the temperature reading now was correct. Second time the altitude readings were all correct.

Seems like there is no algorithm taking in consideration how the start temperature might differ from the rest of the ride

Problem 1: You have to wait for a long time for the ambient temperature to be displayed correctly on the Garmin. Especially while standing still. I really do not have patience to wait 10-20 minutes outside before starting my rides. Imagine waking up 30 minutes earlier in the morning just to stand outside to wait for the Garmin to be able to track my commute to work. No thank you!

Problem 2: Even though I calibrate to 25m (the altitude of my home) before every ride the recorded start altitude is seldom the value I have set manually. It seems like the temperature sensor also influences the offset between manual calibration and the actually recorded altitude.

PS: As mentioned by several others, firmware 4.10 have not solved the problem.

Tried the same tests again. Rode the same round trip twice. The first time I started the Garmin while the temperature displayed was still decreasing after being indoors. The second time I started the tracking right after finishing the first, as the temperature reading now was correct. Second time the altitude readings were all correct.

Seems like there is no algorithm taking in consideration how the start temperature might differ from the rest of the ride

Problem 1: You have to wait for a long time for the ambient temperature to be displayed correctly on the Garmin. Especially while standing still. I really do not have patience to wait 10-20 minutes outside before starting my rides. Imagine waking up 30 minutes earlier in the morning just to stand outside to wait for the Garmin to be able to track my commute to work. No thank you!

Problem 2: Even though I calibrate to 25m (the altitude of my home) before every ride the recorded start altitude is seldom the value I have set manually. It seems like the temperature sensor also influences the offset between manual calibration and the actually recorded altitude.

PS: As mentioned by several others, firmware 4.10 have not solved the problem.

That's a great analysis, however temperature differential indoor vs outdoor is sth that even older units have to deal with. I don't know how or if this issue is tackled in eg. 520 or 810. Perhaps the temp. differential is reduced more quickly or is even more or less the same? I also have 520 and in these colder days I know that elevation data is not spot on (as opposed to summer stable conditions), and I also do not wait more than 5 min to start the ride with "warm" edge 520. So I think you are right that temperature differential affects elevation. I just don't think that this is sth that purely relates to 820. I still think that the basic problem with all elevation issue is that 820 just cannot determine by itself an accurate starting elevation. And as soon as we all start to calibrate elevations we tilt sth else that is related to elevation (ascent, descent, recorded, displayed,...). I think (and hope ;)) that many discrepancies regarding elevation will go away as soon as Garmin resolves the problem of inaccurate starting elevation (of course hoping that an upgrade won't produce another problem :D).

Garmin ????
07/23/2016 reported problem

Garmin ????
07/23/2016 reported problem

Keep Calm and Wait ... and Wait ... and Wait

Elevation issue not fixed in FW5.10

On my device the displayed altitude is still about 40m below the correct value.

I'am wondering whether Garmin isn't willing to fix the problem or their programmers arent't able to fix a consistent error in reading gps altitude.

Is anyone here in the forum who has correct altitude readings? If yes, where do you live?

Cheers,
Bernd

On my device the displayed altitude is still about 40m below the correct value.

I'am wondering whether Garmin isn't willing to fix the problem or their programmers arent't able to fix a consistent error in reading gps altitude.

Is anyone here in the forum who has correct altitude readings? If yes, where do you live?

Cheers,
Bernd

Hi, after this firmware 5.1 if everything before the start I calibrate altitudine there's no issue and it works good but if I don't calibrate it, is like I lived under water but I am 23 meters above sea level! Using 4.1 firmware after calibration altitude went down always as It wasn't calibrated.

Today I have found total ascent 1000 meters and I didn't do any climb. It is the worst device I have ever bought.
I think that programmes aren't able to solve this issue because it is an hardware issue. My old edge 510 was better than this expensive device.