Pronunciation settings define how text-to-speech should say specific words or phrases. They are useful for character names, invented terms, uncommon spellings, and words with different spoken forms.
Use a simple written replacement when that is enough. Use provider-specific pronunciation values only when the target voice requires them and the rendered preview confirms the result.
Choose a preview voice beside New pronunciation. It defaults to the effective narrator. Use the play icon beside any replacement or phoneme in the table or create/edit dialog. One click renders and automatically plays the preview; later clicks replay the completed preview. Capture references such as \1 are omitted from preview audio without changing the saved rule. Preview jobs use the normal rendering workflow, so the selected voice's funding, quota, and provider rules apply. Phoneme support varies by engine and voice.
Drag pronunciation rows into the order you want. The first row runs first, followed by each row beneath it. Put more specific or protective patterns above broader patterns when rules can overlap.
Project pronunciations apply broadly. Book pronunciations can override or supplement the project list for one title.