Voice-led canvas transformation on an iPad
A live tablet demo builds one scene through spoken edits, escalating from a rough sketch to a reflective sunset composition.
Field notes / why it works
The video starts at 0:00 with a spoken object request and a visibly rough pen sketch, then shows a recognizable result by about 0:04. Each later command changes the same composition, expanding from object details to a park at about 0:13 and a reflective final material around 0:32; the mostly continuous tablet view makes cause and effect easy to follow. Four shots across 37.4 seconds leave room to inspect each result without losing the real interaction.
Plate 01 / The format in one breath
A person speaks short creative instructions while drawing on a tablet. The same screen visibly changes from a rough sketch into a finished scene, then transforms again on fresh voice commands. The implicit promise is visible by about 0:04: words and marks can steer a result in real time.
Why it works
- At 0:00, a pen draws a crude chair as a low-register voice names the intended object; the first generated chair is visible around 0:04. This gives a direct input-to-output proof.
- The requests escalate from object to cushion (0:03.6), side table (0:07.5), park (0:10.8), sunset (0:13.2), flowers (0:18.5), and material change (about 0:28–0:32). Each new result raises the scope without resetting the scene.
- The camera holds on the person's hand and tablet for most of the 37.4 s. That continuity makes the interaction legible; only three definite cuts interrupt it.
Plate 02 / Format card
| Platform · aspect · length | X · 16:9, 1280×720 · 37.4 s at 30 fps |
| Pace | 4 shots, 3 definite cuts = 4.8 cuts/min; 4.8–12.8 cuts/min allowing uncertain changes; 8.0 definite picture changes/min; mean shot 9.34 s |
| Script | 77 measured words, 147 measured wpm; 10 short statements averaging 7.7 words; no questions |
| Voice | One conversational adult voice, low register, calm and matter-of-fact |
| On screen | Nearly all live-action, high oblique view of a person using a tablet; product output stays prominent |
| Captions | Small white sentence subtitles with dark outline, centered low; no kinetic highlight |
| Sound | Spoken commands with a low tonal bed about 10 dB under voice; no confirmed accent effects |
| Structure | Initial drawing → object result → scene additions → environment change → detail addition → material-change finale; no spoken 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