The three decisions
- 01Voice casting
An identity has to survive every declared track. Cast for one language and the second gets a different-sounding character, because you are then matching a performance rather than casting a person.
- 02Beat timing
Episodes are timed against dialogue. A line running four seconds in one language may run six in another, and a 90-second episode has nowhere to put the extra two.
- 03Claim substantiation
For branded work, a claim substantiated in one market is not substantiated in another by being translated. The approval does not travel with the words.
Declaring a track you later drop costs a casting session. Adding one you did not declare costs the identity, and there is no version of the second problem that money solves cleanly.
What we are not claiming
Frequently asked questions
Is this dubbing?
Closer to producing the episode in that language than dubbing it afterwards. The track shares the picture, the canon and the story state; the voice, phrasing, captions and review belong to the track, and lip-sync tolerance was chosen with every declared track in mind.
Can I add a language after the season is finished?
Yes, and it costs more than declaring it would have. The system handles the scope change honestly; it cannot undo a casting decision made without that language in the room.
Who signs off on a track?
A native reviewer for that language, as a named approver at their own gate. A Hindi approver cannot sign for Tamil, and asking them to is how localization quality quietly degrades.
Do tracks ever have different edits?
Where a market requires it, and each divergence is recorded. The default is a shared picture, because divergent edits multiply review burden and are the main way a track turns back into a separate project.
What if only one track performs?
That is genuinely useful information and it is part of what the retention gate looks at. It is also one of the few clean comparisons available, because the only variable that changed was the track.
