Camera Extrinsics differe between Registration Output and Metadata(XMP) Output

I have tested many scenes, and the pose of the camera is a little different between Registration Output and Metadata(XMP) Output. The Registration Output, of which the re-projection error is smaller, is frequently better than Metadata.

I used the formula t = - RC  , and found there is some difference between t_reg(exported by Registration) and t_meta(exported by Registration) . 

What is the real meaning of Rotation and Position in Metadata(XMP)?

Hello dear user,
this has already been discussed here: https://support.capturingreality.com/hc/en-us/community/posts/360009527511-Difference-between-the-Metadata-XMP-and-Registration-Files-