Voice cloning

Voice cloning is a technique within audio synthesis that creates a computer-generated copy of a specific person's voice from recorded samples. The output reproduces that person's tone, pitch, and speech patterns, so new spoken audio can be produced in their voice without them recording it.

Why it matters

If you do not know what voice cloning is, you can misjudge what you are hearing in a video. A narration that sounds like a real presenter may be a synthetic copy, which changes how you evaluate the source: an archive clip with original audio carries different evidentiary weight than generated speech laid over footage.

It also affects production decisions. If you build faceless video content, you need to decide whether your voice track is a cloned identity or a neutral synthetic narrator. Cloning a recognizable voice raises consent and platform-policy questions; a generic narrator voice does not carry those risks to the same degree.

An example

A true crime channel producing a Short about a case needs narration over archival news footage. With text-to-speech, the script becomes audio in a stock narrator voice. With voice cloning, the same script could be rendered in the voice of a specific person, for example a reporter whose archive clips appear on screen. ViewMade uses licensed real footage with a media credits file naming each clip's source, and pairs it with narration rather than cloning any individual's voice.

Terms people confuse this with

• Text to speech: converts written text into spoken audio using a general voice, not a specific person's.
• AI voiceover: the broader practice of generating narration with AI, which may or may not use a cloned voice.
• Lip sync: aligns existing audio with mouth movement on screen; it does not create the voice itself.
• Text to video: generates visuals from a prompt; voice cloning concerns only the audio track.