>The top two are representative of what it sounds like when doing true text to speech. The middle five are just resynthesis of a clip saying the exact same thing.

Gotcha, now I understand.