Vox Video Director

Turn a one-line topic into a finished Vox-style paper-collage video: a bold, punchy, narrated explainer/ad where each beat is a torn-paper collage poster that comes alive, with voice-over, optional music and captions. Runs through authenticated Bailian CLI (bl) + local ffmpeg.

The look is the modern editorial paper-collage popularized by Vox explainers and creators like Stav Zilber / rom1trs: hand-cut paper cut-outs, torn edges, tape, halftone dots, newspaper clippings, bold flat color per beat, big cut-out headlines.

The core idea (read this first)

The Vox collage look and the collage motion are two different steps:

1. The look is born in the IMAGE step. Each beat is a finished collage poster made by a text-to-image model. All the collage DNA (torn paper, cut-outs, halftone, bold color, headline text) lives in that image. If the image isn't a rich collage, nothing downstream will save it.
2. The motion is added after. By default an AI video model animates the whole poster (the "living poster" path — simple, automated). For dramatic piece-by-piece assembly you cut the poster into parts and drive them with the local keyframe engine (advanced path).

Everything hinges on the prompts. Before writing any image or video prompt, read references/prompt-guide.md — it has the exact prompt structures that make the difference between "a real Vox collage" and "a moving PowerPoint".

Prerequisites (check, don't skip)

• bl --version — requires Bailian CLI 1.14.3 or newer.
• bl auth status — if unauthenticated, run bl auth login --api-key <key> and stop until login succeeds.
• command -v ffmpeg ffprobe — required for assembly (brew install ffmpeg on macOS).
• python3 -c "import PIL" — Pillow, for captions/watermark overlays.

Standard workflow (topic → film)

This is the default, most-automated path. Every stage is one script, all driven by a single beats.json per project under out/<project>/.

1. Topic → beat map. First read references/beat-layer.md (the story layer) and pick a narrative arc that fits the topic (timeline for history, pas/bab for ads, how_it_works for explainers, man_in_hole for transformations, …). Then write out/<project>/beats.json following that arc. Beat 1 must hook within 3s. Default to one beat = one generated video segment = one shot. Size every segment from its narration, not a fixed template: target 5–10s, use python3 scripts/timing.py out/<project> --estimated, and split any narration that estimates above 10s. A 60s film is usually 7–10 content-sized beats, not six fixed 10s blocks or twelve fixed 5s shots. Vary camera_move across adjacent beats (never repeat; use static on the payoff) and write rich element_motion. Each beat needs narration, title_cn/title_en, scene, bg, feel, hook, and one shots item whose dur may be omitted. This draft with estimated durations is the first mandatory approval gate. Use examples/content-timing.beats.json as the default schema example.

2. Pick the visual style (hybrid — do this BEFORE keyframes). Do not reuse one house style for every topic. Read references/prompt-guide.md (§5 theme presets); pick 3–4 theme presets (styles.THEME_PRESETS: american-retro, swiss-modern, punk-zine, soviet-constructivist, wpa-propaganda, 70s-groovy, chinese-ink, atomic-age, newsprint-editorial) that fit the topic's era/culture/tone — or compose a custom theme by mixing the prompt-guide dimensions (medium/era/palette/type/finish) when none fit. Match the topic, not the language (an English film on Chinese history should look Chinese). A theme bundles the whole LOOK layer (idiom+palette+type+finish+mood+motion). Run a bake-off and let the user pick by eye — AI proposes, the library is the quality floor, the human decides. Set the pick as "theme": python3 scripts/style_bakeoff.py out/<project> american-retro,swiss-modern,punk-zine,atomic-age Set the chosen name as "collage_style" in beats.json (keyframes.py reads it).

3. Voice + measured timing. python3 scripts/audio.py out/<project> Generates one consistent narrator with bl speech synthesize + cosyvoice-v3-flash, then replaces text estimates with measured audio durations plus a short edit tail. New segments clamp to 5–10s. If timing_issues is non-empty, split those narration beats and run audio again before paying for images or videos. Pick voice_id for the topic and language; see references/voices.md. Bailian CLI 1.14.x has no music generation command; provide a local bgm_path or assemble without music.

4. Keyframes (the collage look). python3 scripts/keyframes.py out/<project> Generates one collage poster per beat/shot with bl image generate + qwen-image-3.0, headline text baked in. Compose prompts with the 5-part structure in references/prompt-guide.md. Verify each poster looks like a real layered collage before animating — re-roll cheap ($0.08) here rather than paying to animate a weak image.

5. Motion. python3 scripts/clips.py out/<project> Animates each poster with bl video generate + happyhorse-1.1-i2v. Two independent axes (see references/beat-layer.md §3, tested on our stack): • camera_move — ONE move per shot. Safe/default: {static, push_in, pull_out, pan, tilt, parallax}. Bold/experimental {orbit, dolly_zoom, roll, whip} are available, not banned — they can warp the flat art, so pair with constraints: loose and re-roll. Any custom phrase also passes through. • element_motion — where the energy lives; AI writes it per beat to fit that scene (not a template). Make it RICH (several elements moving) — be bold. A hero element flying across the frame (paper bird/plane/coins) is a great occasional punch on a key beat, not every shot (a flyer in every frame reads as a formula). motion_style = amplitude calm | punchy | max (the theme sets a default). constraints = strict (default: defect guards on — flat-2D, one-way, no-morph; best for clean text-heavy explainers) or loose (let the model explore 3D/bold moves; re-roll the misses). Headline text is hard-protected only on shots that have a title (detail shots without a headline are free to go wild). The Bailian default is happyhorse-1.1-i2v; validate real-person and brand use against the selected model's current policy before a full run. Aspect routing (styles.resolve_video_aspect, second approval gate): clips.py resolves doc["aspect"] against the chosen video_model's own supported ratios — exact match wins; Omni is 16:9/9:16 only, Kling reference-to-video adds 1:1, Kling image-to-video/video-edit and Seedance just follow the input/ratio param. When there's no exact match it picks the nearest ratio but stops and asks you to confirm (set "aspect_approx_confirmed": true once you have) rather than silently reframing the film — every clip in one run shares the same resolved aspect so the finished film is never mixed.

6. Assemble. python3 scripts/assemble.py out/<project> ffmpeg: normalize + concat all shots, lay the single narration ducked under the music, burn captions timed per beat, add the watermark. Output out/<project>/final.mp4.

7. Verify. You can't read an mp4 directly — extract frames to jpg and look: ffmpeg -ss <t> -i final.mp4 -vf "scale=640:-1,format=yuvj420p" -frames:v 1 f.jpg

Cadence — content decides duration

• Default to one narration beat and one generated clip per segment. Keep each segment in the model-safe 5–10s range.
• Estimate before generation from text length; after TTS, treat measured narration_dur as authoritative and round up to the next 0.5s with a ~0.45s edit tail.
• Video APIs receive the next whole second (for example 6.5s → request 7s); assembly trims the returned clip to the exact edit duration, so narration is never cut.
• Short text still gets 5s so the idea can land. Long text gets up to 10s. If narration plus tail exceeds 10s, split it at a sentence or clause boundary; never speed-read or silently stretch a clip past the model limit.
• A 60s film usually lands at 7–10 segments of mixed lengths. Avoid identical durations across the whole film unless the source genuinely has equal-length beats.
• Use multiple shots inside one beat only as an explicit editorial choice. Legacy multi-shot projects remain supported, but new content-driven projects should split the narration into separate beats so each generated clip still has its own 5–10s content unit.

A-roll mode (talking-head → collage)

The standard workflow above is B-roll: a topic becomes AI-generated collage posters that get animated. A-roll is the reverse case — the user already has a real recorded talking-head video (a presenter speaking to camera) and wants it itself turned into the collage look, keeping their actual performance (face, lip movement, gestures) intact. There is no poster to generate; the "keyframe" is the presenter's own footage. Use A-roll when the user gives you a video file of themselves/a presenter talking, not a topic to write from scratch.

1. Transcribe + auto-segment. python3 scripts/asr_beats.py <project_dir> <source.mp4> Runs bl speech recognize (fun-asr) on the source audio and cuts it into beats at sentence-ending punctuation or natural pause gaps (never exceeding ~9.5s, under Omni/Kling video-edit's 10s per-call cap). Writes beats.json with each beat's start/end/text — this is the same mandatory approval gate as the B-roll beat map: review it, set "theme" (run style_bakeoff.py the same way — the presenter's segment works fine as the bake-off source), and optionally fill in a content_beats string per beat (a sticker/stamp idea to layer in) before generating anything.

2. Generate. python3 scripts/aroll_clips.py <project_dir> [only_ids] Cuts each beat's time range out of the source, uploads it, and re-styles it with a photographic paper-cutout sticker treatment on the presenter — her real likeness, lip movement, eye-line and gestures follow the source frame-for-frame; only the silhouette edge and the world around her are paper-collage. Default model is happyhorse-1.0-video-edit through bl video edit (set via video_model/video_model_fallback in beats.json). Never ask the model to redraw or halftone-texture the face itself — that gets rejected regardless of how the prompt is worded (tried both a strong and a softened phrasing; both failed). Uses the same aspect-routing confirm gate as clips.py.

3. Assemble. python3 scripts/aroll_assemble.py <project_dir> Muxes each generated clip with the original beat segment's own audio (never whatever audio the video model produced) so lip-sync is guaranteed regardless of which model handled that beat, normalizes every beat to one canvas, and concats into final.mp4.

C-roll mode (one photo → collage)

The third input modality — "cutout roll". A-roll re-styles a talking-head VIDEO; B-roll generates everything from a topic; C-roll takes a single still PHOTO (a selfie, an avatar card, a product shot) and anchors it inside the collage world: the subject is cut out as a PHOTOGRAPHIC sticker — never redrawn — and per-beat posters are generated around it with an image-EDIT model, then animated through the normal clip stage. Use C-roll when the user gives you one photo and a topic: a personal explainer fronted by their own face, or a collage ad built around a real product shot (validated on both, 2026-07-17).

1. Beat map. Same as B-roll (references/beat-layer.md, same approval gate), plus the C-roll fields in beats.json: "mode": "croll", "anchor_photo", "croll_subject" (portrait | product), and subject_wardrobe (portrait — lock the outfit or the paper-doll body drifts) or subject_desc (product). Set "title": false on shots — C-roll posters carry no headline; text belongs to captions. If there is no separate script, transcribe/derive narration first and let the audio's ASR timestamps define the beats (audio-first, like A-roll — not text-first like B-roll).

2. Anchored keyframes. python3 scripts/croll_keyframes.py <project_dir> Uploads the photo once and generates one anchored poster per shot via qwen-image-3.0 through bl image edit. Portraits get a photographic face + illustrated paper-doll body; products get a pixel-faithful sticker with label typography intact. Prompt rules that are baked in (all three cost a re-run to learn): poses/expressions go to the BODY only — asking for a wink redraws the face; halftone must be scoped to the background or it bleeds onto skin; portrait clothing must be locked explicitly. The script also writes anchor_freeze into beats.json.

3. Voice timing + animate + assemble. Standard audio.py → clips.py → assemble.py. clips.py injects the anchor_freeze guard into every motion prompt — without it the video stage can re-letter a product label (observed: "PARFUM" → "PAREUM") or re-time a face. For narration in the subject's own voice, set voice.clone_ref (see Voice + music above); derive stamp/snap-zoom timing from the narration's ASR word timestamps (asr_beats.py works on any audio, not just A-roll footage).

beats.json schema

{
  "project": "my-film", "topic": "...", "language": "en",
  "aspect": "9:16",                       // 16:9 | 9:16 | 1:1 | 3:4
  "style": "collage",
  "provider": "bailian_cli",              // default; invokes authenticated `bl` child processes
  "theme": "american-retro",              // THEME_PRESET (styles.THEME_PRESETS) — the LOOK layer
  "arc": "timeline",                      // narrative arc (beat-layer.md) — the STORY skeleton
  "video_model": "happyhorse-1.1-i2v",
  "image_model": "qwen-image-3.0",
  "image_resolution": "1k",               // 1k (default) | 2k | 4k
  "video_resolution": "720p",             // 720p (default); Seedance also 480p/1080p (Omni is 720p-only)
  "motion_style": "punchy",               // amplitude: calm | punchy | max (theme sets a default)
  "constraints": "strict",                // strict = defect guards on | loose = let AI explore + re-roll
  "timing_mode": "content",               // default: text estimate, then measured TTS duration
  "timing": {"min_segment": 5, "max_segment": 10, "tail": 0.45},
  "voice": {"voice_id": "longtian_v3", "language": "zh", "speed": 1.0}, // see references/voices.md
                                          // + optional "clone_ref": "path/to/sample.mp3" (clone that voice via seed-audio)
                                          //   and "persona": "YouTube tutorial creator" (delivery style for cloned VO)
  "bgm_path": "path/to/instrumental.mp3", // optional; omit to assemble without music
  "mix": {"music": 0.6, "voice": 1.25},   // audio balance — optional; these are the defaults (BGM ducks under the VO)
  "caption_style": "white",               // white (default: clean white subtitle) | paper (cream cut-out collage look)
  "captions": true,                       // false = no burned-in captions (deliver clean, subtitle in post)
  "watermark": "Made By 阿里云百炼CLI",
  "mode": "croll",                        // C-roll only — plus the four fields below
  "anchor_photo": "path/to/photo.png",    // C-roll: the still to anchor (person or product)
  "croll_subject": "portrait",            // C-roll: portrait | product
  "subject_wardrobe": "a cream knitted sweater and charcoal trousers",  // C-roll portrait: outfit lock
  "subject_desc": "the perfume bottle",   // C-roll product: short noun phrase for the sticker
  "beats": [
    {
      "id": 1, "title_cn": "", "title_en": "BEFORE MONEY",
      "bg": "earthy clay tan", "feel": "ancient, humble", "hook": "surprising_stat",
      "narration": "For most of history, there was no money...",
      "shots": [
        // Omit dur: timing.py/audio.py sets a content-driven 5-10s duration.
        {"id": "a", "title": true, "shot_size": "WIDE", "camera_move": "push_in",
         "scene": "...wide establishing collage...",
         "element_motion": "traders gesture, goat bobs, a paper bird crosses, coins scatter"}
      ]
    }
  ]
}

theme+arc set the two big layers; element_motion per shot is the energy (make it rich — see below). motion/collage_style/era are still read for back-compat.

Bailian model selection

Use bl model list when you need to verify or override an ID. Defaults for this edition:

Job	Model	Note
Keyframe / collage poster	qwen-image-3.0	called through bl image generate
Anchored image edit	qwen-image-3.0	called through bl image edit
Animate / image-to-video	happyhorse-1.1-i2v	called through bl video generate
Talking-head restyle	happyhorse-1.0-video-edit	called through bl video edit
Narration	cosyvoice-v3-flash	called through bl speech synthesize
ASR	fun-asr	called through bl speech recognize
Music	local file	set bgm_path; optional

See references/models-and-gotchas.md for the full model-choice reasoning and every API / ffmpeg gotcha (auth header, curl downloads, no-libass captions, content blocks, etc.). Read it before debugging any failure — most failures are already documented there.

Backends are pluggable. Every generation call goes through scripts/provider.py. bailian_cli is the default and starts parallel bl child processes without exposing an API key to the project. atlas_cloud remains as an explicit compatibility backend. Failed CLI jobs may retry; running CLI jobs are never duplicated on a stall because that could create duplicate charges.

Advanced: element-level motion collage

The standard path animates the whole poster (great, automated, "living poster"). For the dramatic pieces-fly-in-and-assemble motion collage (à la cr7v2), or to animate real people with full control and zero content filters, cut each poster into independent elements and drive them with the local keyframe engine (no video model needed).

Read references/local-engine.md. In short: extract_elements.py (crop + background-removal

• residue/erase cleanup) → motion.py (Layer + keyframes, fly_in/slap/drop/pop_settle easings, procedural confetti/starburst, camera zoom+shake+whip, frame render). Pieces fly back to their original positions on a blurred-placeholder backdrop, so the assembled frame reconstructs the original poster.

Editions

• Auto edition (this skill): topic in, film out, through Bailian CLI.
• Manual prompt-pack: if the user cannot run Bailian CLI, just produce the beat map + the per-beat image prompts + the per-clip motion prompts + the narration script for them to paste into any generator. The creative engine (the prompts) is identical.