This is an Auditory Illusion
What is the person saying?
The audio is of a guy saying "bah"
The video is of a guy saying "gah"
When a person focuses on his face they commonly hear "thah" or "dah"
When a person looks away they hear "bah" perfectly.
This is called the "McGurk effect"
http://en.wikiped ia.org/wiki/McGurk_e ffect