The alignement of the mask is of paramount importance to keep the
drive in the mask into the upper part, to produce some vignetting
beyond row 235.
At each setup, one must check that the vignetting of the upper rows
is properly achieved.
It can easily be tested, just overexpose on the sky and watch how
the counts change when you move the mask towards more negative values,
something like -500 enc relative.
At the moment the three objectives are aligned such that approx. 5
rows at the upper edge of the detector are vignetted, in order to
prevent the MUX from being illuminated.
NB: it's not enough to aling the slits and take a fixed offset for
the detector mask encoder position: indeed, after aligning the slits,
the mask will be perpendicular to the array, but then, the objectives
are adjusted such that the mask is imaged nicely parallel to the upper
border of the array at ~ row 235