Hello, you all.

Since my last firmware update, I have a rather important issue with altitude measurement.

I went out today for some tests of the camera and did two flights at a  very short distance from my home point (max home distance about 150')  and at relatively low altitude (no more than about 50').

If altitude appeared to be well estimated at the begining of both  flights, it decreased with time and reached negative values while I was  hovering or moving slightly forward or backward at the same level.

When I landed approximatively at my home point (take off), altitude  measurement was as low as -156 feet (see the 1st flight log hereunder,  10 min to the end).

The "drift" in altitude measurement begins at roughly 6m 40s and continues till landing, increasing regularly.

Strange enough, the last data after switching off the motors returns to the right value (zero feet).

What can explain these errors and how to correct it ?

I can not fly Litchi missions with such huge errors.

The first flight analysis :
http://www.phantomhelp.com/LogViewer/AMM77FUT5YFX65EJRU2B/

The second one, less regular but also erratic in altitude :
http://www.phantomhelp.com/LogViewer/T3CPPWKBOE6FOL429BTU/

Thank you in advance for your answers...

[Edit : I'd like to add that the Mavic did not climb or went down while  the altitude measurement was drifting. It remained absolutely stable,  hovering or moving at the same high]                                                

I haven't run tests like that but it appears that the altitude may be based on barometric pressure and I have seen variances like what you report.  I also noticed on my first flight, just a bit over 14,000 ft msl, (Above the summit of Mt. Democrat) the Mavic reported a max altitude of over 15,000 feet which was about 1,000 too high. This is more typical of a barometric pressure reading than a GPS altitude discrepancy but I really am just speculating on this.

Yes, you can have barometric pressure changes during a flight but if you observe the changes in altitude measurements during the two flights realized at the same location with less than 1h interval, you'll see that these changes towards low values are almost identical and that they begin at almost the same moment during the flights.

It seems to me very unlikely that natural variations in atmospheric pressure could produce this.

Anyway, I have contacted DJI support and I'm waiting for an answer.