Music-Responsive Cybernetic Portrait
A one-take capture of a robotic 3D portrait responding to music inside a live control interface.
Starter prompt / by Refs
The creator shared no short prompt. This is ours for making a film like it with Claude.
Model a cybernetic robot head in Blender with Python, split into movable parts (visor, temple plates, crest, jaw, neck servos), and load it into a native macOS app written in Swift with SceneKit. The app plays {AUDIO_FILE}, analyzes bass, mids and transients in real time, and maps them to each part's rotation, extension and light pulses. Add a dark side panel with sliders and presets for every mechanism, then screen-record a 30-second 4K demo of the head performing to the track.Field notes / why it works
At 0:00 the finished metallic portrait and its live sliders share the frame, making both output and controls immediately clear. The head pivots to profile by 0:02.75, returns toward front around 0:07 and 0:14, then turns again near 0:24.6; those changes sustain a 28.1-second shot without cuts. Keeping the controls visible makes the motion read as an interactive demonstration.
Plate 01 / The format in one breath
A polished 3D object performs inside a dark technical interface while its control panel stays visible. Frame one shows both the result and the machinery: the viewer can inspect whether the audio really drives the object. The promise is visual proof of a working interactive experiment, supplied by the image and the post text rather than a spoken introduction.
Why it works
- At 0:00, a large metallic head and a dense slider panel establish the output and its controls in one frame. The viewer understands the experiment immediately.
- Between 0:00 and 2:75, the head rotates from front to profile, crown pieces open, and gold accents pulse; there is motion inside a single shot, with no edit needed to imply capability.
- By about 0:14 the head returns to a front view with another light state, then changes again through 0:24.6. The 28.1-second hold makes repeated responses observable, rather than presenting one reveal.
Plate 02 / Format card
| Platform · aspect · length | X · 1216×720 (about 1.69:1) · 28.08 s · 50 fps |
| Pace | 1 continuous shot · 0 cuts/min · 0 detected scene or graphic steps/min · 28.08 s shot length; object motion is continuous |
| Script | Four machine-transcribed words, but the isolated “Music” at 0:17.5 has very low confidence. No explanatory narration; the post caption carries the premise. |
| Voice | Brief voice-like exclamation near 0:26, likely part of the track (inferred); no presenter or voiceover |
| On screen | 100% desktop/app capture: roughly three-quarters 3D viewport, one-quarter control panel; no person |
| Captions | None. All text is native interface labels and tiny technical callouts. |
| Sound | Music-led track, roughly 121 BPM (low-confidence estimate); −16.6 LUFS integrated, −0.7 dBFS true peak |
| Structure | Four observable movement phases; the demonstration starts at 0:00, continues through 0:28, and ends without an 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 / The hook, frame by frame (first 3 seconds)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