An APERTURE offset is made in a coordinate system fixed in the instrument rotator, so its value in RA and Dec changes as a function of rotator angle. This is done to facilitate moving an object between points on the detector itself, rather than defining offsets on the sky for which OFFSET should be used.
There are three apertures: NOMINAL, A and B:
The reference point need not coincide with the axis of the instrument rotator, and the system knows the displacement between them, if the software is setup correctly. Thus, an image placed on the reference point will remain stationary when the rotator is turned.
Aperture offsets are applied as pointing corrections: the displayed position of the source does not change when they are introduced (except briefly, as the telescope moves). This is because aperture offsets are used to move the same object to a different place on the detector, rather than a different object to the same place on the detector (for which operation, an RA/Dec OFFSET should be used). There is no reason why both kinds of displacement should not be used simultaneously. For example, if the crosswire is used as reference point for SNAFU, the following sequence can be useful: (1) slew to a bright reference star, (2) centre it on the reference point, (3) RA/Dec OFFSET to a faint galaxy, and (4) beamswitch A to move that galaxy onto the spectrograph slit.