I figured that I would tell people about something new I created. It is called the SeeSong engine. It works using two costumes or backgrounds (mouth open and closed), the timer, a song file, and a list. The list contains times that tell when to switch costumes/backgrounds. If this is done while a sound is played, it can give the effect of singing. I have a project to demonstrate this here: http://scratch.mit.edu/projects/ahawesome/366661. I've found a use for this in my other project, Musical You. It guides the user through the process of importing the required mouth open/closed pictures and makes them a singer. It is available here: http://scratch.mit.edu/projects/ahawesome/366651. The reason I am telling people about this is in case anyone wants to make a project based off of the code. I think that people will find it very interesting and a source for inspiration.
Offline