For my singing snowman, I had several different faces defined for the various combinations. The mouth positions were set the same for each face, but the "outline" would add the Hat or the Scarf or the buttons or similar. For the "Faces" effect, I could then just select which of those definitions I wanted to use.