---
name: filmscene-shots
description: "Write and run one part of a film-scene remake on MiniMax H3 — your measured scene notes as shots with cut_at, setup image, camera, action and acting from the original, dialogue with exact times and off-screen lines; H3 Max reference-to-video (two setup images + Audio 1) or H3 Max Turbo image-to-video (first/last setup + target_audio); checks of cuts and sync; the comparison with the original. Part of the filmscene reference set."
---

These creative references describe source experiments with external tools. Use your own supported account and tools; this download installs no automation. Settings and timings are experimental observations, not current FlashBoards availability. Use only media and voices you have permission to use.

# filmscene-shots · one part, one generation

> **Alternative external route:** Seedance 2.5 reference-to-video (`bytedance/seedance-2.5/reference-to-video`, references as @Image1… / @Audio1…, 4–30 s, native sound) was preferred in the source experiments. The measured side-by-side comparison was a product commercial, not every format; compare it with the H3 method below for your material.

## Prompt (canon v2 structure; fill the slots from the person's cast)

Use your own scene notes prepared by `filmscene-corpus`. The prompt retains the source term "corpus move" for a camera move recorded in those notes; no private corpus or prompt runner is supplied or required.

```
speakers: {character A}: S1, {character B}: S2
{ output: one continuous film scene, D seconds, 21:9, N shots with hard cuts exactly at the given times; Audio 1 is exactly D seconds…,
  references: { Image 1: setup A — every {character A} shot looks exactly like it…, Image 2: setup B — every {character B} shot…, Audio 1: the whole dialogue… },
  characters: { {character A}: {look: the person facing the camera in Image 1…, voice: {character A}'s voice of Audio 1 (name it by what tells the two voices apart, e.g. the man's / the woman's / the lower voice)…}, {character B}: {…} },
  environment: { location…, light…, sound: only Audio 1; no music },
  style: live-action 35 mm film, natural and restrained…,
  shots: [ { shot, cut_at, from, to, image, camera (setup + "static" from the corpus move), action: BEGINNING… THEN… END…, acting,
             dialogue: [ { speaker: S1, on_screen: true, when: 00:02.140–00:03.200, line: <d>[English] "…"</d> } ] } ],
  rules: [only the person facing the camera speaks…; an off-screen line is heard from the person seen from behind and the face on screen keeps its mouth closed; no cut to the speaker inside that shot; one full-screen shot at a time; no phones] }
```

- Looks come from the setup images only (blind) — never write a character's face or clothes into `look`.
- Times are **the voice track's measured word times** (Scribe), not the subtitle times.
- An off-screen line: `on_screen: false` + the rule that the listener's mouth stays closed. Measured: H3 Max kept the listener's mouth closed and the line in place.

## Requests

- **H3 Max reference-to-video**: `reference_image_urls` [setup A, setup B], `reference_audio_urls` [voice.mp3], `duration` D, `resolution 480P`, `aspect_ratio 21:9`, `prompt_expansion_mode disabled`.
- **H3 Max Turbo image-to-video**: `image_url` first setup, `end_image_url` last setup, `target_audio_url` voice.mp3, same D/480P/disabled; the prompt says "the given first frame / last frame / the given audio" instead of Image/Audio numbers. Output 992 × 416.

## Measured (example run, 480P)

| Part | Engine | Cuts (corpus → got) | Lines |
|---|---|---|---|
| 1 · A–C–A, 15 s | Max | 8.212 → 8.79, 11.747 → 11.50 | speaker 1 0.00 s; speaker 2 −0.95 s (re-voiced by the model, still in their shot) |
| 2 · C–A, 12.9 s | Max | 4.849 → 4.75 | all 4 lines 0.00 s, off-screen line kept |
| 2 · C–A, 12.9 s | Turbo | 4.849 → 4.67 | our audio as is (corr 1.00) |

## Checks after every generation

- Cuts: `ffmpeg -vf "select='gt(scene,0.25)',showinfo"`.
- Sync per line: cross-correlate the generation's audio with `voice.wav` on each line's span (shift and corr).
- Contact sheet 1 fps → faces, the off-screen moment.
- Make your own comparison file (e.g. `cmp.mp4`): the original on top, the generation below (960 × 400 each), the generation's audio + the film room tone. Review cuts, sync and faces yourself before continuing; source-author approval is not required. `voice.wav` above means your own prepared voice track, not a supplied file.

## Example run (reference only — never reuse its files, prompts, faces, voices or frames for a person)

*Heat* (1995) units 996–998 with the demo characters: `speakers: dex: S1, mara: S2`; Image 1 = setup A (every Dex shot), Image 2 = the second camera setup (C in the measured scene; every Mara shot); Dex's voice = "the man's voice of Audio 1". In the table speaker 1 = Dex, speaker 2 = Mara; the off-screen check kept Dex's mouth closed.

The table's corpus times are the source experiment's measured scene times. No example files are delivered or required; compare your own measurements.