Stopping ends current output without disabling future feedback. Quiet mode lets you work temporarily without speech. Restart after correcting a backend problem or changing the selected synthesiser, not as an assumed cure for every missing integration.
Stop current main and notification speech with tts-stop.
Toggle current-buffer quiet mode; C-u toggles it globally.
Stop and restart the selected speech server.
Speak adapter and version information.
Server selection, repeated-pattern cleanup, text splitting, language selection, and direct engine controls have specialist Help and reference entries. For missing speech, follow Recovering From No Speech.