Voice & Audio
This is the voice StreamBuddy uses for its own replies, AI chat, game announcements, and ad reads. Reading viewers' messages aloud is a separate feature (see Chat TTS).
Windows Voices (SAPI)
Built into Windows, robotic/classic TTS, no setup needed. Audio plays through your default speakers; OBS picks it up via desktop audio capture. No OBS overlay required.
Microsoft Neural Voices
High quality, natural-sounding. Uses Microsoft Edge's speech API (requires internet) and plays inside the OBS browser source as its own audio track. Examples: en-US-ChristopherNeural, en-US-JennyNeural.
Note: In OBS, right-click the browser source, go to Properties, check "Control audio via OBS". Skip this and the voice synthesizes fine but you won't hear it.
Speed
Recommended speech speed is 1.0x–1.2x. Above 1.3x sounds rushed; below 0.9x feels unnatural live.
Voice Volume
One 0–100% slider that applies everywhere StreamBuddy speaks — AI replies, Chat TTS, Stream Ads, custom commands and greetings, timed posts, shoutouts, and game announcements. It is separate from the Browser source volume in OBS, so you can use both. Soundboard clips keep their own per-clip volume.
Tip: AI replies, Chat TTS, game announcements, and ads all share one playback queue, so they never talk over each other, they play one after another.