One Portrait, Infinite Character Variations
A music-led character-consistency launch film that turns one reference into fast cinematic variations, then proves the workflow in a brief UI reveal.
Field notes / why it works
A plain reference portrait holds for 2.7 seconds before rapid cinematic results make the consistency promise legible without narration; eight cuts land in the first 10 seconds. New people around 0:25, an unexpected cat sequence at 0:35–0:48, and a compact UI glimpse around 0:49 renew attention while escalating proof. Large, sparse claims and small source-image insets let viewers verify the transformation even when muted.
Plate 01 / The format in one breath
A silent-reading, music-led product announcement demonstrates a single-input creation tool through escalating proof. It opens on a plain reference portrait, turns one identity into many polished scenes, repeats the experiment on people and an animal, briefly shows the product interface, then resolves to a spare brand card.
Why it works
- The first 2.7 seconds establish a fixed framed portrait. Rapidly replacing the portrait, then revealing styled versions of a character around 0:05–0:12, makes the consistency claim visible before the UI appears.
- The first 10 seconds contain eight cuts; the film then gives the viewer longer proof views, including a nearly four-second hold around 0:08–0:12.
- New subjects at about 0:25 and 0:35 re-open the test. A cat DJ, skateboarder, runway model and cook from about 0:41–0:48 show the range without adding explanatory narration.
Plate 02 / Format card
| Element | Reference |
|---|---|
| Platform · aspect · length | X · 1282×720, 16:9 · 60.8 s · 24 fps |
| Pace | 41 detected shots, 40 cuts; 39.4–54.2 cuts/min allowing uncertain changes; mean shot 1.49 s, median 1.38 s; 41.4 definite picture changes/min |
| Script | No verified spoken narration. About five short, large on-screen claims plus the closing brand line. The machine transcript after 0:30 appears to capture vocals in the music, not product speech (inferred). |
| Voice | No presenter or narrator; music may contain sung or rhythmic vocals from about 0:30 (inferred). |
| On screen | Roughly 80% cinematic generated-result montage, 12% reference/input or interface, 8% title and end cards (visual estimate from the sheets). |
| Captions | None. Large editorial claims replace subtitles: black on off-white, or white over darker imagery. |
| Sound | Continuous, loud, high-energy music-like bed; rough tempo 179 BPM or half-time 89.5 BPM. No clear cut-synced effects measured. |
| Structure | Input → human style range → one-image claim → more people → animal range → interface → clean logo/payoff. No direct in-video CTA. |
Pro / 13 plates + recipe
The first plates are open. The full dissection is Pro.
Get the timed shots, type and sound specs, and the runnable recipe.
Open the full blueprint- Plate 03 / Hook, frame by frame (0:00–0:03)Pro
- Plate 04 / Structure (beat sheet)Pro
- Plate 05 / Shot grammarPro
- Plate 06 / Visual style guidePro
- Plate 07 / Voice and deliveryPro
- Plate 08 / Sound designPro
- Plate 09 / PackagingPro
- Plate 10 / Script templatePro
- Plate 11 / Production planPro
- Plate 12 / PromptsPro
- Plate 13 / Edit recipePro
- Plate 14 / QA checklistPro
- Plate 15 / Make it yoursPro
- Recipe / runnable promptPro