First decisions on language, dialect and voices

Decide the language of every spoken part now, before a single line is written, and flag anything that changes the plan (your internal shot plan, made at Stage 04) or the budget.

The decisions (decided), each marked "from the brief" or "inferred"; an inferred one is confirmed at the meeting:

1. Language versions are locked by the brief. A new version is the client's decision.

2. For each spoken or sung part, decide its kind: voiceover, on-screen dialogue, song or jingle, a real filmed interview, or on-screen text only.

3. Lock market, dialect and sub-dialect (see The brief and research → Dialect rules). "White" Arabic is made by casting plus pan-Arab wording, so ask which audience it faces.

4. Name the sign-off for each dialect now, and ask for any reviewer the project needs.

5. Who appears is a casting fact (real talent, staff, an ambassador). Ask who; there are no rights questions.

A first read on voices [best available, not yet validated]: an AI voice per dialect, a native voice artist as the fallback (casting itself happens at Stage 05, Anchors). Flag now:

• Sa'idi: no AI Sa'idi voice has been found, so a native Sa'idi actor is likely.

• Children's voices: unproven in Arabic; an adult native actor playing the child, or a child's own read.

• A real person's voice, or an existing brand voice: a professional clone from 30 minutes or more of that person's clean speech, on a higher ElevenLabs plan; or the person records.

• English principal lines: recorded first (decided). Seedance's own voice. Option: use it when a principal line is in English or another language Seedance speaks well, and you choose it for the project [proven]. It is chosen per project, never a default, and never for Arabic (decided): Arabic principal lines stay on recorded voices. The client approves the voice from a sample made with one fixed voice description (see Speaking shots and lip sync → English lines and Seedance's own voice).

• A song, a jingle, a sung voiceover, or a film the brief wants cut to its music: flag it music-first now. Its Suno track is made and locked with the anchors, and a real singer is booked for the finals, which changes the schedule and the budget (see Music → The music-first route). Sung Arabic is spelled as the dialect is spoken, with vowel marks as the voice rules set them (see Voice and Arabic dialects → Spelling for the engine).

• Real filmed footage or interviews: flag them now, because they set the film master's frame rate and add sound cleaning and a paper edit (see Editing → Filmed footage and interviews).

Record it as one table in the project file:

Sources

• This Guide's own rules.