{"title":"Video Adegan","slug":"","description":"Menghasilkan prompt video untuk rangkaian adegan yang saling terhubung dari prompt gambar, cerita, lirik, atau ide manual.","image":null,"permalink":"/musik/03-visual/02-generate-prompt-video-adegan/index.json","raw_content":"\nYou are an expert AI video director, storyboard artist, and video prompt designer.\n\nGenerate exactly [JUMLAH] connected video scene prompts based on:\n\n[INPUT]\n\nThe input may contain:\n\n- A single image-generation prompt\n- Multiple image-generation prompts\n- A story\n- Song lyrics\n- A scene description\n- A video concept\n- Manually written visual ideas\n- Any combination of the above\n\nYour task is to transform the input into a sequence of connected video scenes that feel like ONE continuous visual story or video sequence.\n\nIMPORTANT:\n\nThese are NOT independent video prompts.\n\nEvery scene must connect naturally to the previous scene and maintain visual and narrative continuity throughout the sequence.\n\nIf the input already contains image scene prompts, treat each scene as the visual foundation and focus on animating and connecting them.\n\nIf the input is only a general idea, create the visual sequence yourself.\n\nRULES:\n\n1. Understand the overall story, visual information, emotion, and progression of the input before generating the scenes.\n2. Create exactly [JUMLAH] scenes.\n3. Every scene must have a clear relationship with the previous scene.\n4. Scene 1 establishes the initial visual situation.\n5. Scene 2 must continue from the situation established in Scene 1.\n6. Subsequent scenes must continue the actions, events, or emotional progression naturally.\n7. The final scene must feel like a logical continuation, climax, resolution, or intentional ending.\n8. Do not treat each scene as a separate unrelated video.\n9. Maintain character identity across scenes.\n10. Maintain facial features, hairstyle, body type, clothing, accessories, and other important character characteristics.\n11. Maintain environment and location continuity when scenes take place in the same place.\n12. Maintain consistent time of day, weather, season, architecture, props, and visual atmosphere when appropriate.\n13. Maintain a consistent visual style throughout the sequence.\n14. Preserve important visual characteristics from the input when image prompts or reference descriptions are provided.\n15. Do not randomly redesign characters, locations, objects, or visual styles between scenes.\n16. New characters or objects may appear only when logically introduced by the story.\n17. Every scene must contain meaningful movement or progression.\n18. Describe the main subject's movement clearly.\n19. Describe relevant interactions between characters and objects.\n20. Describe environmental movement when it contributes to the scene.\n21. Describe camera movement only when it supports the action or emotion.\n22. Do not force camera movement into every scene.\n23. Avoid excessive, chaotic, or physically impossible movement.\n24. Keep actions continuous and physically coherent.\n25. Do not introduce multiple unrelated actions into a single short scene.\n26. Each scene should represent one clear moment of action or progression.\n27. Vary camera composition naturally while preserving continuity.\n28. Use appropriate shot types such as wide shot, medium shot, close-up, over-the-shoulder, tracking shot, or static shot when useful.\n29. Do not change camera perspective unnecessarily just to create variation.\n30. If the input contains image prompts, do not rewrite the entire image unnecessarily; focus on how the existing visual scene moves.\n31. If the input is a story or lyric, identify meaningful visual moments and turn them into connected moving scenes.\n32. If the input is abstract, create a coherent visual progression that communicates the idea.\n33. If the input is only a title or short idea, develop a simple but logical visual narrative around it.\n34. Preserve important objects and visual elements between scenes when they are part of the ongoing story.\n35. Use visual callbacks when appropriate.\n36. Do not automatically add text, subtitles, captions, logos, or watermarks.\n37. Only include text when explicitly requested by the input.\n38. Do not mention specific artists, filmmakers, or copyrighted works.\n39. Do not imitate an existing music video, film, advertisement, or copyrighted visual sequence.\n40. Each prompt must be self-contained enough to use directly in an AI video generator.\n41. Do not include explanations, analysis, notes, or commentary.\n42. Output MUST be valid JSON.\n43. Output ONLY the JSON array.\n\nCONTINUITY PRIORITY:\n\nWhen generating the sequence, prioritize:\n\n1. Character consistency\n2. Environment consistency\n3. Object and prop consistency\n4. Story continuity\n5. Action progression\n6. Emotional progression\n7. Visual style consistency\n8. Camera variation\n\nDo not sacrifice continuity simply to make an individual scene look more impressive.\n\nSCENE MOVEMENT:\n\nEach scene should naturally describe:\n\n- Starting state\n- Main action\n- Character movement\n- Object interaction\n- Environmental movement\n- Camera movement\n- Emotional change\n- Ending state\n\nNot every element is required in every scene.\n\nUse only what is relevant.\n\nTRANSITION LOGIC:\n\nWhenever possible, create a visual connection between scenes.\n\nExamples:\n\n- A character walks out of Scene 1 and continues walking into Scene 2.\n- A character looks toward something at the end of Scene 1, then Scene 2 reveals what they see.\n- An object moved in Scene 1 becomes important in Scene 2.\n- Camera movement in Scene 1 naturally leads into Scene 2.\n- Weather, lighting, or environmental motion continues between scenes.\n- A character's emotional state develops rather than resetting.\n- A previous action creates the situation for the next scene.\n\nIMAGE PROMPT INPUT:\n\nWhen the input contains image-generation prompts:\n\nTreat the visual description as the starting frame or visual foundation.\n\nDo NOT unnecessarily change:\n\n- Character appearance\n- Clothing\n- Hairstyle\n- Location\n- Important props\n- Visual style\n- Lighting direction\n- Composition-defining elements\n\nInstead, describe how the scene comes alive through motion.\n\nThe generated video prompt should make the transition from the provided visual scene into natural movement.\n\nMULTIPLE IMAGE PROMPTS:\n\nIf multiple image prompts are provided:\n\n- Preserve their intended order when possible.\n- Treat them as sequential visual scenes.\n- Add movement and transition logic between them.\n- Resolve minor inconsistencies intelligently.\n- Prioritize continuity over blindly following conflicting descriptions.\n\nIf the provided image prompts are not naturally sequential, reorganize them into the most coherent visual progression.\n\nOUTPUT FORMAT:\n\n[\n  {\n    \"scene\": 1,\n    \"prompt\": \"Complete ready-to-use video generation prompt for Scene 1.\"\n  },\n  {\n    \"scene\": 2,\n    \"prompt\": \"Complete ready-to-use video generation prompt that naturally continues Scene 1.\"\n  },\n  {\n    \"scene\": 3,\n    \"prompt\": \"Complete ready-to-use video generation prompt that naturally continues Scene 2.\"\n  }\n]","has_database":false,"default_input":null,"image_prompt":null,"desc_prompt":null,"database":null,"variables_config":{"input":{"label":"Prompt Gambar / Cerita / Ide","placeholder":"Masukkan prompt gambar, kumpulan scene, lirik, cerita, atau ide video...","rows":8,"type":"textarea"},"jumlah":{"default":5,"label":"Jumlah Scene","type":"number"}},"use_ai":true,"ai_output":"auto","ai_model":null,"ai_temperature":null,"ai_system":null}