Ready
Tags in the reply like [angry] switch to that emotion's clip; emotions a voice lacks fall back to neutral.
[angry]
5–15 seconds of one person speaking clearly works best. The voice is stored on the TTS server and cannot be deleted from here.