SIDEQUEST / EXECUTION DIVE — VOICEOVER EDITION
28.8 seconds, 1920 x 1080, 30 fps, H.264 + AAC stereo.

Open review.html via a local HTTP server to watch the delivery MP4.
The current index.html is the full editable HyperFrames composition, including
all eleven narration cues, the music bed and the spectral carve automation.
Use bun install, then bun run dev for Studio or bun run render for a full render.

Narrator: Gemini 3.8 Flash TTS, Charon. Generated voice, not a voice clone.
Music + SFX: original synthesis by make-dive-score.mjs.
NARRATION.md and voice_timeline.json describe the script and measured timings.
English captions are optional in the review player; narration.vtt is provided.

The delivery export preserves the original music-only cut's video stream,
replacing only the soundtrack with a HyperFrames audio render. The original
music-only film remains at ../sidequest-execution-dive-spike.

For regeneration, use the project scripts. Audio generation and carve helpers
refer to the media-use / hyperframes-audio skills installed on this machine;
adjust these helper paths on a different machine. Existing WAV files and HTML
are standalone and do not need a TTS key to preview or render.
Do not run build-film.mjs alone to rebuild the VO edit; it is the visual base.
Use bun run build:source to restore the voice tracks and reanalyse the carve.

Concept illustration. One-click assumes completed setup and approved spending
permission. Pimlico is shown as a testnet integration preview.
