Natural, human-sounding voices, recorded for every sitting at a calm, unhurried pace. New sittings are usually ready to hear within a few minutes of arriving.
Device voice (fallback only)
Used only when a sitting has no natural recording yet. Device voices can sound robotic.