I am well aware that restraints are very sensitive, and can greatly impact a static study. But I was wondering whether the same holds true for frequency studies, or that it is even more important to approximate them accurately (e.g. not just fixed).