Najlepiej jednak chyba przetestować samemu.
Baader podaje: If you have f.e. a coma corrector, field flattener, focal reducer, focal extender (to a lower degree due to concave surface), or in extreme cases a whole lens group for more complex field corrections a few centimeters in front of the filter it could be useful to flip the filter against the rule from above (thus having the arrow pointing away from the telescope). Cause in such cases the likelihood of reflections from the sensor could be lesser then fort- and back- reflections from such glass-surfaces. If in doubt, it helps to make some test images from a star field with bright stars, using the filter in both ways for comparison.