The decision record

01.What is recorded

Translation: which rendering carries a joke, a story or a conversational turn; the slang kept as said, the name kept, the reference swapped, what the culture avoids, the profanity level. Your creative directors review these today: three translations per line, each with its reasoning, pick one or edit it.

Audio (in the works): what to translate at all, what stays in the original language (a chant, a yell, a phrase), which sounds are kept as they were. You choose what is dubbed and what is left alone.

Visual (in the works): who on screen is saying it, the text in the frame (a phone screen, a street sign), the expressions and gestures. You pick a different face, translate or keep each sign, or hold an expression as shot.

02.Read the record

The record as an API resource is in the works. One page per request, every decision with its options and what was chosen.

{
  "dub_id": "dub_31xk9m",
  "lang": "en",
  "decisions": [
    {
      "id": "dec_7h2k0a",
      "type": "translation",
      "kind": "rendering",
      "segment": "seg_12",
      "start": 12.48,
      "end": 15.91,
      "speaker": "spk_1",
      "source": "사과하려고 사과 사 왔어.",
      "options": [
        { "id": "opt_1", "text": "Apples. I'm sorry, right down to the core.", "intent": "the pun kept", "emotion": "playful", "reasoning": "The line rests on 사과 meaning both apology and apple, and bringing apples to say sorry is a real custom; the English keeps the fruit so the bag in the next shot still lands." },
        { "id": "opt_2", "text": "I came to say sorry. I brought apples.", "intent": "the meaning only", "emotion": "sincere", "reasoning": "Drops the pun; the apology reads straight." },
        { "id": "opt_3", "text": "An apple a day keeps the silent treatment away.", "intent": "a new joke", "emotion": "playful", "reasoning": "A different gag that fits the apples on screen." }
      ],
      "chosen": "opt_1",
      "chosen_by": "model",
      "status": "open",
      "memory": null
    },
    {
      "id": "dec_7h2k0b",
      "type": "audio",
      "kind": "keep_original",
      "start": 41.20,
      "end": 44.05,
      "speaker": null,
      "source": "crowd chant",
      "options": [
        { "id": "opt_1", "text": "keep in Korean", "reasoning": "A stadium chant; the crowd is the scene, not a speaker." },
        { "id": "opt_2", "text": "translate", "reasoning": "The words carry the next line's joke." }
      ],
      "chosen": "opt_1",
      "chosen_by": "model",
      "status": "open",
      "memory": null
    },
    {
      "id": "dec_7h2k0c",
      "type": "visual",
      "kind": "on_screen_text",
      "start": 58.00,
      "end": 61.50,
      "speaker": null,
      "source": "phone screen: 내일 봐",
      "options": [
        { "id": "opt_1", "text": "See you tomorrow", "reasoning": "The message is read on camera; the terms you already chose apply." },
        { "id": "opt_2", "text": "keep as shot", "reasoning": "Leave the screen in Korean." }
      ],
      "chosen": "opt_1",
      "chosen_by": "model",
      "status": "open",
      "memory": null
    }
  ],
  "has_more": false
}

Every delivered dub will also carry decisions.json (in the works): the full record per language, with what was chosen and by whom (Outputs & subtitles).

03.The catalog of kinds

04.Lifecycle

With review=manual every decision is open at in_review, and POST /render locks the record. With review=autothe record is still written with the model's choices (chosen_by: "model"), readable after render.

To change a delivered dub, PATCH the decision, then POST /render again; only the languages the decision touches render again. The dub.decisions_ready webhook (in the works) fires with dub.ready_for_review, carrying open and by_type counts (Webhooks).

05.Override

curl -X PATCH https://api.thefamiliarlab.com/v1/dubs/dub_31xk9m/decisions/dec_7h2k0a \
  -H "Authorization: Bearer $FAMILIAR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "choose": "opt_2", "remember": "creator" }'

The override endpoint is in the works. Send exactly one choice field, plus an optional remember.

The response is the decision with chosen_by: "you" and status: "settled". PATCH is valid at in_review and complete (then render again).

Batch

{
  "decisions": [
    { "id": "dec_7h2k0a", "choose": "opt_2" },
    { "id": "dec_7h2k0b", "choose": "opt_1" }
  ]
}

06.Settings on the dub

Planned request fields on POST /v1/dubs (in the works); in multipart, nested fields are written dotted, as metadata.title already is (Create a dub).

ask anywhere requires review=manual; otherwise the request returns 422 decision_setting_requires_review. The dub waits at in_reviewuntil every open decision is settled or you render, which accepts the model's choice on anything still open.

07.The review screen

Creative directors work on translation decisions today in the review step. In the works: the review screen gains a decisions rail, a timeline of the episode with a marker per decision, filtered by TRANSLATION, AUDIO or VISUAL.

Each card will show the moment, the options with the reasoning and the model's pick; one tap picks another, types your own, keeps the original or remembers the choice for this creator. It is the same record the API returns, so a choice made on the screen shows in the API and the other way round.

08.Access