[avandonk]

As usual it is all a compromise. Ideally you want perfect polar alignment and the guidescope guiding on the center of your image. If your mount has both scopes, imaging and guiding at right angles to the polar axis that is your declination axis then there will be no drift or image rotation.

In the real world this is not true. You will get drift and rotation which gets worse as the alignment errors get larger.

Misalignment of guidescope with imaging scope will tend to give image rotation and this gets really important if you are doing widefields. This is especially true even at short focal lengths. The center star (guidestar) will look fine and the outer ones will be tangentially smeared due to rotation.

Drift is caused by slight polar misalignment or so called cone error or both.

A small amount of drift and or rotation as long as it is not noticeable on a single exposure can be beneficial. This occurs when you stack a series of exposures as the noise will be in slightly different places on each exposure. The noise will then not add but tend to cancel out especially if you median stack. It works even better with INCR as the difference is bigger between exposures. Or you can stack the ODD and EVEN frames seperately and simply then average the two stacks.

Hope this helps.

Bert