Seedance 2.5 demo: full prompt behind a 30-second photorealistic travel-fashion film
AravPhi · x · 2026-09-04
The author shares the complete prompt used to generate a 30-second, photorealistic travel-fashion film with Seedance 2.5.
Key techniques in the prompt:
- Identity locking: opens by fixing the same woman's facial identity, hair, makeup, body proportions and jewelry, explicitly banning identity drift, face morphing and plastic CGI skin.
- Cinematic parameters: 24fps, 180° shutter, natural motion blur, subtle film grain; only 24mm environmental, 35/50mm portrait and 85mm detail lenses with physically achievable camera moves.
- Timed wardrobe control: three outfits bound precisely to 0-7s (New York), 7-16s (farmers market), 16-30s (California coast), switching only at 7s and 16s.
- Realism constraint list: demands correct walking mechanics, hands, hair and clothing physics while forbidding duplicated people, teleporting, warped architecture, random logos and watermarks.
- Audio policy: location sound only — traffic, footsteps, ocean, laughter — no dialogue, score or artificial SFX.
The prompt is a reusable template for controlling character consistency, storyboard pacing and realism in video generation.
Related event: Seedance 2.5 Generates 30-Second Travel Fashion Film from a Single Prompt(2 posts)→
More from Multimodal
- PrunaAI's P-Video-Edit lands on Runware: text-instruction video editing at $0.045/s — aziz4ai · 2026-09-04
- Practical guide to extending videos and tagging inputs with Gemini Omni Flash — fofrAI · 2026-09-04
- Omni 1.1 adds video references: 3-second clips ensure character consistency in AI video — fofrAI · 2026-09-04
- AI-made Resident Evil x Psycho horror parody, part one — No-Invite8044 · 2026-09-04
- MiniMax H3 ecosystem boom: cinematic LoRAs, 360° video, and one-take chaining tools — optimisticalish · 2026-09-04
- Gemini-based agentic video understanding experiment goes viral, repo coming soon — MarioLucic_ · 2026-09-04