Speech synthesis turns written language or structured directions into an audio performance. A system determines how words are pronounced and shapes timing, emphasis and vocal qualities to produce a spoken result.
Uses include narration, accessibility tools, conversational interfaces and draft dialogue. Direction can influence the delivery, but pronunciation, emotional expression and pacing may still need correction. A convincing voice does not establish the truth of what it says.
ELI5
If you type a sentence and a computer reads it aloud, you are using speech synthesis. Some systems can also follow directions such as speaking slowly or sounding excited.
For example, you could create a draft narration for a short film. Listening back lets you catch a wrongly pronounced name or a pause that makes the sentence sound strange.
Is speech synthesis the same as speech recognition?
No. Speech synthesis produces spoken audio, while speech recognition converts incoming speech into text or another representation.
Can a synthesized voice follow performance directions?
Some systems accept directions about delivery, pacing or emotion. How reliably those directions work varies, so the resulting audio still needs listening and review.

