De carte de RPG à scène d'événement anime
Transformez une carte de village de RPG en vue de dessus en une scène d'événement anime à hauteur des yeux mettant en scène des aventuriers fantastiques, idéale pour les cinématiques de visual novels et le concept art de jeux vidéo.
Prompt
À l'aide de l'image de référence fournie, transformez la carte de village RPG en vue de dessus en une illustration d'événement soignée dans un style anime, vue à hauteur d'yeux humaine. Conservez le même emplacement de village et les repères d'agencement : le puits en pierre central, le réseau de sentiers, les haies, les maisons en bois et l'étroit canal d'eau sur la gauche. Convertissez les 2 petits personnages sprites près du puits en 2 personnages de fantasy grandeur nature au premier plan : un mage aux cheveux argentés vêtu d'une robe violette tenant un bâton, et une elfe blonde en tenue d'aventurière verte et brune, tous deux penchés et regardant dans le puits. Ajoutez une ambiance cinématique de JRPG avec une douce lumière du jour, un rendu pictural détaillé, un line art net et une légère profondeur de champ. Préservez l'impression que cette scène se déroule sur la même place du village, mais enrichissez-la d'une perspective naturelle, de davantage de détails environnementaux et de 5 villageois en arrière-plan : 1 homme coupé au bord de l'extrême gauche, 2 petites silhouettes debout au loin sur le chemin central, et 2 habitants en train de discuter près des bâtiments sur la droite.
Guide d’utilisation
Ce guide d’utilisation est actuellement disponible en anglais.
Converts a top-down RPG town map into an eye-level anime event illustration while keeping the same landmarks in the same places. Best when you already have a map asset and want a cutscene that players will recognize as the same square.
When to use
Visual-novel event CGs, RPG cutscene stills, pitch art for a tile-based game, or a key visual for a town hub. The composition needs one central anchor object in the foreground with both characters interacting with it, and the map's left and right features must stay on the same sides.
What the existing cover image shows
The existing cover shows a round stone well in the foreground with a silver-haired mage in a purple robe holding a blue-orb staff on the left and a blonde elf in a green cloak and brown leather on the right, both bending toward the well with startled expressions. Cobblestone paving, trimmed hedges, iron lamp posts, timber-and-stone houses, a large tree and a narrow canal at the far left all read clearly. The extras only partly match the brief: a man cropped at the left edge is there, the center path holds one distant figure rather than two, a couple talks at right under an awning, and a pink-haired woman is cut by the right edge. Because the source map is not shown, layout fidelity cannot be judged from the cover alone, and although water is visible inside the well there is no hint of what the two are reacting to.
Suggested setup
- Aspect ratio
- 16:9 · A 16:9 landscape frame leaves room for the well, both characters and the village behind them; the listing marks 1:1 but the cover is widescreen. Ask for a human eye-level camera with the well slightly below center, the characters cut at the thighs by the rim, and a gentle depth of field that keeps the houses readable.
- Reference images
- Required: the top-down map image you want converted. Use your own tile map or assets you are licensed to reinterpret; a map lifted from a commercial game will carry its art into the result.
What to customize
- Layout anchors kept from the mapA lighthouse at the far end, a wooden pier network, a stone breakwater on the right, fish-market stalls, a cliff path on the left
- List four to six landmarks with their side of the frame; with fewer the model invents a generic village, with more it starts rearranging them.
- The two foreground charactersA red-haired knight in blue-and-silver armor and a dark-haired healer in a white-and-gold robe
- Describe hair, outfit colors and one held item per character, plus a single shared action toward the anchor object so they are not posed independently.
- Background extrasAn old fisherman cropped at the left edge, two children on the distant pier, two merchants talking by the market stalls
- Give exact counts and positions, keep them small, and state that no other people appear; extras easily multiply or steal focus.
View a complete adaptation
Harbor map to anime dawn scene
The same map-to-illustration conversion applied to a harbor town, with a knight and a healer examining a washed-up net at the pier and five positioned townspeople.
This adaptation has not been generated. The existing cover image comes from the original prompt collection and is not a result of this adaptation.
Using the provided reference image, transform the top-down RPG harbor-town map into a polished anime-style event illustration at human eye level. Keep the same location and layout cues: the lighthouse at the far end, the wooden pier network, the fish-market stalls, the cliff path on the left, and the stone breakwater on the right. Convert the 2 small sprite characters on the pier into 2 full-size fantasy characters in the foreground: a red-haired knight in blue-and-silver armor holding a lantern, and a dark-haired healer in a white-and-gold robe, both crouching over a torn fishing net full of glowing shells. Cinematic JRPG feel: pale dawn light, painterly rendering, clean line art, gentle depth of field. Keep it recognizably the same harbor square, with natural perspective, richer environment, and 5 background townspeople: 1 old fisherman cropped at the far left edge, 2 small children on the distant pier, and 2 merchants talking by the market stalls on the right. No other people.
Troubleshooting
- The result ignores the map and produces a generic fantasy village
- Name four to six specific landmarks with their left, center or right placement, repeat 'same town, same layout, eye-level' once, and keep the reference map as the only attachment.
- The crowd count drifts: extra villagers, missing ones, or a second copy of a hero
- State the number of extras and their positions as a short numbered list, keep them small and distant, and end the prompt with 'no other people'.
Editorial notes and generation record
AI-assisted editorial analysis of the source prompt and existing illustration. Settings are starting suggestions; the new adaptation is not a model test.
No new illustration was generated for this guide. The observations refer to the prompt's existing cover image from its source collection; its generation details are not recorded.
Détails du prompt
- Catégorie
- Assets de jeu
- Modèle
- GPT IMAGE 2
- Coût par image
- 1K : 43 crédits · 2K : 86 crédits · 4K : 143 crédits
- Format
- 1:1
- Auteur
- @無量光秀＠アートゥルデザイン
- Licence
- CC BY 4.0
- Ajouté le
- Mis à jour le