A Comparative Study of Self-supervised Speech Representation Based Voice Conversion