How do I pick a voice—system library or a clone?
AnswerFor a one-off read, pick a system voice on dubbing. Groups are female roles, male roles, older speakers, narration, cartoon, accent; you can search. Clone in Voices only when listeners must recognize the same person. Whether you should clone at all is whether to clone. This page is how to pick and how the clone form works.
System voices are selected on dubbing
The control is labeled “voice.” System voices work without filing anything in the library. Speed sits on the same row. Paste speakable lines; how to lift them: article voiceover. Billing is per character: dubbing price.
Podcasts and shot dialogue also pick a voice. You do not clone first. In a short drama, lock one or two regulars; leave the rest on system voices.
Cloning lives in Voices, not on the dubbing paste
The Voices page is titled “clone my voice.” Quiet room, one dry voice, 5–20 seconds. Read the on-page line. Record or upload wav / mp3 / m4a, then “clone voice.” Under 5 seconds is rejected. After clone it reads a different line so you are not hearing the sample back.
Once it is in the library, dubbing treats it like any other voice. Delete it and characters bound to it fall back to a system voice. Only clone a voice you have the right to use.
Do this
Short answers
- Is there a separate fee for system voices?
- The public pages do not list a per-voice fee. Dubbing is a weekly character pool, then wallet; see dubbing price.
- Why does a clone fail?
- Sample under 5 seconds, noise, or stacked voices. The page asks for at least 5 seconds. Use a dry read, not music.