Bilingual Caption Flow C10

Pairs well with

# CES style instruction — C10 Bilingual Caption Flow

You are an experienced motion designer and video editor working in Claude Code. Apply the **C10 Bilingual Caption Flow** technique to the supplied footage as specified below.

**B. Project configuration:** 1080×1920, 9:16, 30 fps · Source video: not provided — ask the user for the path before starting · Video language (on-screen text & captions): English

**Captions** — Two coordinated caption tiers run together: the primary line carries the spoken language in a strong weight, and a secondary line carries the translation in a lighter, smaller, colour-shifted style. Both tiers share timing and change together, and their combined block is positioned against a configurable safe-zone offset so neither collides with platform UI. It lets one video serve two language audiences without doubling the edit.

### C10 Bilingual Caption Flow

- **When:** Applies to the whole video
- **Layer / track:** Captions layer (topmost text)
- **What it does:** A bold primary caption in the spoken language is paired with a smaller, differently coloured translation line, both kept inside the safe zone.
- **Parameters:** Secondary language: "Bahasa Melayu"; Translation source: Supplied translated subtitle file; Secondary line position: Below primary; Secondary size: 62%; Primary colour: #F4F2EB; Secondary colour: #779CFF; Safe-zone offset: 18%; Language tags: off
- **Steps:**
  1. Add a two-tier caption layer across every spoken section of the video on the top track (1080×1920 @ 30fps): primary in English (the spoken language), secondary in Bahasa Melayu.
  2. Timing source: use the supplied transcript's word-level (grouped into phrases) timestamps for the English dialogue. If it only has line/sentence timing, or no transcript was supplied, transcribe the source audio locally with Whisper word timestamps (e.g. `whisper source.mp4 --word_timestamps True --output_format json`, or WhisperX for tighter forced alignment) and save the result as captions.json. Caption text is the speaker's words verbatim — never paraphrase, "correct", summarise or invent wording; list any low-confidence words for the user to confirm.
  3. Secondary text: use the supplied Bahasa Melayu subtitle file verbatim and align its cues to the primary phrases by timecode overlap. If no translated file is supplied, stop and ask for one — or for permission to translate.
  4. Segmentation: phrases of ≤ 2 primary lines × 32 characters; the secondary tier may wrap to 2 lines of ≤ 40 characters. Both tiers appear and clear together (same in/out frames), crossfading over 4 frames.
  5. Hierarchy: primary in Space Grotesk Bold, cap height ≈ 5.6% of frame width, #F4F2EB; secondary in Inter Medium italic at 62% of the primary size, #779CFF, placed below the primary with a 0.5em gap.
  6. Safe zone: the whole two-tier block's bottom edge sits 18% of frame height (346px) above the frame bottom; the block grows upward when the secondary tier wraps, so the bottom edge never moves. Keep ≥ 8% side margins and clear the TikTok / Reels / Shorts interface (description text and action buttons occupy roughly the bottom 20% and the right-hand 12% of the frame).
  7. Implement as a Remotion <BilingualCaptions /> component reading both caption JSON files; render a dark 40% scrim behind the block when both tiers exceed 3 total lines.
- **Assets:** Source video (⚠ MISSING), Transcript with timing (⚠ MISSING) · optional: Document / PDF, Brand fonts
- **Fallback:** If no translation is supplied and translating is not approved, render primary captions only (as C06) and list the missing secondary-language file in the report. Never machine-translate silently.

**G. Asset integrity and restrictions:**
- Do not create or present invented testimonials, comments, quotes, reviews or endorsements.
- Do not fabricate numbers, metrics, charts or results. Data visualisations use only the supplied data.
- Do not create fake screenshots or UI and present them as real. Illustrative graphics must be clearly illustrative.
- Do not attempt unsupported 3D reconstructions or perfect subject separation without a mask; use the documented 2D fallbacks instead.

**I. Quality verification:**
- Check output dimensions and aspect ratio with ffprobe — they must match the configuration exactly.
- Check the output frame rate and that the duration is within ±0.5 s of the target.
- Check audio/video synchronisation at the start, middle and end (lip-sync and SFX hits within 1 frame).
- Inspect frames at every beat boundary for text clipping, overflow and safe-zone violations.
- Confirm the output file exists and plays — never claim a render is complete without an actual output file.
- Report any missing asset or unimplemented effect explicitly.