We're working on an update that will allow you to choose a unique voice for each paragraph in your content.

You can use this to simulate conversations, or just have multiple voices in the same audio file.