"Correctness" of Visual vs. V for unusual objects?