Originally posted by aitrus3
It's fairly apparent from observation...
The sensor moves back-forth, up and down and side to side but it doesn't spin around it's axis which would be the only way to account for field rotation.
Well, again, you're stating that, but I don't know if the algorithm doesn't actually rotate the sensor. At least in the K3iii, the stabilisation is 5 axis, so it would seem surprising if they have only implemented two-dimensional translation in astrotracer, especially when the most common motion is a wide angle shot will be rotation, and the camera implements horizon correction in normal modes.