Stop asking Midjourney for 'cinematic': a full painterly thriller-director system prompt
CurieuxExplorer · x · 2026-09-29
A complete system prompt for Midjourney V8.2 is circulating: it casts the model as a 'painterly animated thriller director,' so users stop stacking words like 'cinematic' and instead let a director-level prompt steer the whole 'film.'
Key points:
- Targets painterly animated film stills covering portraits, dialogue, action, environments, and connected stories, with urban suspense as the default direction, adaptable to other genres
- Builds dimensional faces, bodies, and environments from painted color planes, visible brushwork, and selective drawn details — painting with preserved volume, not photos with paint filters, plastic renders, or flat vectors
- Works with or without references and moodboards
As the quoter puts it: 'Most prompts describe a picture. A system prompt directs the film.'
More from Multimodal
- ElevenLabs v4 puts direction in the script: [whispers], [laughs] and natural-language delivery notes — xiaohu · 2026-09-29
- Dev builds auto-storyboard app linking ComfyUI to local Gemma, generates 80 scenes at once — Flaky_Comedian2012 · 2026-09-29
- Dev vibe-codes a node to fix garbled text in AI-generated speech bubbles — arturor1990 · 2026-09-29
- Early 'Opus 5.5' animation test drops: Into the ANTVERSE — ChrisGPT · 2026-09-29
- On DGX Spark, bf16 beats int8 convrot: H3 video gen 272s vs 287s in real tests — dtdisapointingresult · 2026-09-29
- Kling 4.0 hands-on: 10-bit HDR, 30s single-shot generation and 15 reference inputs for film-grade AI video — APPSO · 2026-09-29