Dear pix4d team,
Some time ago we processed an aerial data set in the Netherlands with pix4d 2.0.89.
Recently, we acquired a couple of GCPs in the area to verify the data sets accuracy properly.
Subsequently, we processed the data set again (in 3.1.22), and it turned out that there was a rather large offset between the GCPs and the aerial data before adjustment.
The GCPs, however, perfectly fit to the data set processed earlier with the older version (2.0.89),
We were wondering if there were some changes regarding the height reference, as we think this might be the reason for this issue.
In 2.0.89, the data has been processed using the Bessel ellipsoid in the RD New system. Of course, we used the same reference again with the newer version.
Thank you for your help! Please let me know if you need further information.