Image Prompts for Comic Graphic Novel

a movie poster layout for "{argument name="title" default="TITLE"}" with an expansive white background acting as negative space, featuring reference image. high-quality 2D anime illustration medium, cel-shaded with soft 3D-like rendering, using a stark white baseline palette contrasted with the colors of the reference image. Korean manhwa. note: it requires at least 1 image as reference. note 2: add this in the prompt to remove the poster credits: negative: billing block or credit block

Using REFERENCE_0 as the manga-name/storyboard base, convert the rough layout into a polished full-color vertical manga page. Keep the overall 4-panel structure: 1 wide top panel, 2 diagonal split middle panels, and 1 wide bottom panel, but render it as finished shoujo-style manga art with clean black panel borders instead of editor guide boxes. Goal: Replace the placeholder character boxes with fully illustrated characters and keep the storyboard’s intended dialogue beats. Page header: Add a simple printed header with {argument name="chapter title" default="夜食"} at the top left and {argument name="page number" default="2/4"} at the top right. Characters and panel actions: - Panel 1: Render the “acha” placeholder as a pretty young woman in a warm kitchen/dining setting, holding a small bowl of tofu topped with garnish while speaking. - Panel 2 left: Render the “devil Fab” placeholder as a long silver-haired beautiful man with black horns, small bat wings, a dark outfit, and an annoyed expression with arms crossed, against a darker purple devilish background. - Panel 2 right: Render the “angel Fab” placeholder as the same silver-haired man in an angel form, smiling with hands clasped, white wings, halo, flowers, sparkles, and a warm glowing background. - Panel 3 bottom: Render the young woman again at the left, holding a small red-capped spice bottle labeled 七味; include a dramatic spiky burst callout in the center and the remaining speech bubbles on the right. Dialogue and text: Use the following Japanese text exactly where it belongs in the visible balloons and callouts: {argument name="all dialogue text" default="じゃ…お豆腐にめんつゆかけて…… / ちっ…冷奴は正直おいしいので何も言えません… / せめてネギとおかかは載せましょう それくらいの悪は許されます / まぁ…！夜食界の優等生ですよ… / えらいです… / 七味も…かける…！ / くっ…今夜は天使側の勝ちですか… / それは悪ではなく彩りです"}. Style: Finished clean anime/manga illustration, soft painterly coloring, attractive character designs, detailed hair and clothing, subtle screentone-like backgrounds, crisp speech balloons, vertical Japanese lettering, professional web manga page quality. Constraints: Preserve the reference’s main composition and dialogue order, use exactly 4 panels, exactly 8 visible dialogue/callout balloons plus the spice bottle label, no extra characters, no watermark except a small centered handle at the bottom: {argument name="bottom handle" default="@_sagyo"}.

Goal: Create a vertical black-and-white Japanese manga page about an absurd office crisis involving a hamster employee, drawn in detailed screentone line art. Canvas: Tall vertical comic page, approximately 9:16, with a clean white border and four horizontal panels stacked from top to bottom. Use crisp manga ink lines, grayscale shading, halftone screentones, speed lines, exaggerated reaction faces, and dramatic speech bubbles. No color. Setting: A modern open-plan office with desks, computer monitors, binders, filing cabinets, partitions, paperwork, pens, and fluorescent corporate lighting. Keep the environment consistent across panels. Main subjects: Include exactly 3 recurring characters: 1 young Japanese male office worker in a dark suit, white shirt, striped tie, short black hair, seated at a desk with a monitor and keyboard; 1 angry middle-aged male manager in a dark suit and glasses, seated or standing at a rear desk, with a stern face that escalates into shouting; 1 fluffy hamster employee sitting on a tiny office chair at a miniature desk, with a tiny monitor, keyboard, papers, and office supplies. The hamster should look cute but serious, like a competent coworker. Layout and panel action: Panel 1: Calm office scene. The young office worker types at his computer on the right foreground, looking slightly concerned or distracted. The hamster sits at its tiny workstation on the left foreground. The manager is in the background behind another desk, watching sternly. No large dialogue in this panel. Panel 2: The manager suddenly shouts from the background with a huge jagged speech bubble and intense speed lines. The young worker flinches in surprise with a small sweat mark and sound effect. The hamster remains at its tiny desk. Use exactly 1 large jagged speech bubble in this panel containing {argument name="manager first outburst" default="今すぐヤ〇トの集荷呼べ！！今日中！！熊本に支援物資を送るから！！"}. Panel 3: The young worker turns toward the manager, anxious and sweating, while the hamster types rapidly at the tiny keyboard. The manager is furious in the background. Use exactly 3 speech bubbles in this panel: 1 large jagged bubble from the manager containing {argument name="manager shock line" default="はぁ！？"}; 1 rounded bubble from the young worker containing {argument name="worker delivery reply" default="く、熊本ですか……？配送ストップしてると思いますが……"}; 1 rounded bubble near the hamster containing {argument name="hamster delivery report" default="はい、どの配送会社も熊本への配送は全域ストップしてます"}. Add small typing sound effects near the hamster. Panel 4: Dramatic dark background replaces the office, creating a surreal despair scene. A clock on the wall shows about 5:30. The hamster is at the bottom left behind its tiny keyboard with blank glowing eyes, and the young worker is at the bottom right with blank exhausted eyes, sweat drops, and a defeated expression. Use exactly 1 large jagged speech bubble on the left containing {argument name="manager final outburst" default="そんなの無理を通してこい！！話せばどっかしら受け取ってくれるだろ！！片っ端から当たれ！！"}. Visible repeated elements to include: exactly 4 panels; exactly 3 recurring characters; exactly 1 tiny hamster workstation; exactly 1 normal foreground workstation for the young worker; exactly 1 rear manager desk; exactly 1 clock in the final panel; exactly 3 binders grouped on the young worker’s desk edge in the office panels. Visual style: Authentic monochrome seinen manga, detailed office backgrounds, clean panel borders, expressive faces, dramatic jagged balloons, hand-drawn Japanese lettering, realistic perspective, high contrast in the final panel. Avoid color, avoid Western comic styling, avoid extra panels, avoid extra characters, avoid replacing the hamster with any other animal.

Create a full-body minimalist studio portrait of {argument name="character description" default="a young woman with long wavy brown hair in a high ponytail, warm medium skin tone, hoop earrings, a black T-shirt, olive cargo pants, and black-and-white sneakers"} jumping energetically in midair against a clean off-white studio wall and pale concrete floor. Show a precise vertical split down the center of the body: the left half is realistic color photography with natural soft window light, visible fabric texture, realistic hair movement, and a raised left fist; the right half is transformed into a black-and-white hand-drawn manga/comic sketch version of the same pose, matching the identity, clothing, proportions, hair, and motion exactly. The character’s legs are bent as if leaping, one sneaker kicked back on the photo side and the other descending on the sketch side. Add exactly 14 comic-style drawn background elements only on the sketch side: 3 outlined stars, 2 cartoon smoke clouds, 2 lightning-bolt marks, 5 long speed-line clusters, 1 impact burst at the floor under the foot, and 1 group of small ink dots. Keep the background minimal and uncluttered, with the split line perfectly aligned through the face, torso, pants, and shoes. Use high-resolution editorial lighting, playful action energy, crisp realism on the left, expressive ink hatching on the right, no text, no watermark, no extra people.

Goal: Create a monochrome Japanese manga page with an accidental surreal-comedy tone: an elegant, melancholy young man in a late-night private study has a dramatic conversation with a cute hamster as if it were a serious noir mystery scene. Canvas: Vertical manga page, black-and-white with subtle cool blue-gray tinting in the man's hair and eyes, high-resolution screentone shading, crisp black gutters, printed manga texture. Use exactly 8 panels arranged in a dense page layout. Layout: Panel 1 is a full-width cinematic establishing shot at the top, showing a traditional Japanese-style study at midnight with shoji window grids, rainy dark garden visible outside, hydrangeas, stacked books and drawers, a glowing rectangular lantern, and the man seated at a desk smoking. Panel 2 is a left medium close-up of a hamster peeking over a desk edge. Panel 3 is a right close-up of the man's face, holding a smoking pipe or long cigarette holder near his lips. Panel 4 is a left medium shot of the man reclining with the hamster in his hand while smoke curls upward. Panel 5 is a narrow central hamster reaction panel. Panel 6 is a right close-up of the man gently holding the hamster near his face. Panel 7 is a full-width extreme close-up of the man's eyes and hair, separated by a thin horizontal gutter. Panel 8 is a full-width dramatic final panel with a black background, floating white petals, butterflies, smoke wisps, and the man smiling behind an open folding fan. Subject details: The main character is {argument name="character name" default="an elegant mysterious young man"}, early twenties, androgynous bishonen features, pale skin, long tousled {argument name="hair color" default="dark blue-black"} hair falling over one eye, sharp red-gray eyes, long dangling earrings, layered dark kimono or formal robe, refined hands, calm teasing expression. The hamster is {argument name="hamster description" default="a tiny round gray-and-white hamster with glossy black eyes, small paws, and a confused innocent expression"}. Emphasize the absurd contrast between ultra-serious romantic manga framing and the hamster's cute presence. Text content: Add vertical Japanese manga narration and speech bubbles throughout, using this exact overall dialogue/narration set where it fits: {argument name="Japanese dialogue" default="深夜の私設書庫にて。 / こんな時間まで付き合ってしまったね。 / おや、君はまだ起きていたのかい？ / ふふ、その驚いた顔も可愛いねぇ。 / き、君のせいじゃないのか!? / 君のせい、というわけでもないよ？ / ほら、こうして… / けれど、君がいてくれると夜更けも退屈しない。 / ……さて、そろそろ本題に入ろうか。 / 今夜の議題は、君の処遇だよ。 / 夜は、まだ始まったばかり。"}. Include small handwritten sound effects near the hamster and smoke, such as tiny squeaks and rustling marks, but keep them secondary. Visual style: Detailed shojo-seinen hybrid manga illustration, elegant linework, glossy hair rendering, dramatic close-ups, cinematic lighting from the lantern and moonlit windows, fine screentones, high contrast black backgrounds, delicate smoke trails, petals and butterflies in the final panel. The mood should feel ornate, mysterious, and unintentionally funny. Constraints: Use exactly 8 panels, exactly one human character, exactly one hamster, one folding fan in the final panel, one glowing lantern in the study, multiple speech bubbles but no modern UI elements, no color except subtle blue-gray tint and faint red eyes, no watermark.

Goal: Create a colorful full-page Japanese manga-style promotional comic poster for {argument name="platform name" default="AniFusion"}, a creator platform that helps users make comics, illustrations, videos, drawings, and manage assets. The mascot is {argument name="character name" default="Ai-kun"}, a cheerful anime cat-boy guide. Canvas: Vertical A4 poster, white outer margin, thick black comic panel borders, high-energy shonen manga advertisement layout, bright pastel accents, clean print-ready composition. Main character details: A cute young anime cat-boy with fluffy brown hair, large brown cat ears, a fluffy tail, round black glasses, warm brown eyes, small whisker marks, a white hoodie under blue denim overalls, red collar with a gold bell, and a yellow hair clip reading “AI”. He appears repeatedly as the guide mascot, smiling, winking, pointing, waving, and presenting tools. Layout and counted panels: Use exactly 9 main comic/content panels plus one right-side vertical text strip and one bottom banner. Panel 1, top-left large panel: close-up of Ai-kun winking and pointing a stylus toward the viewer, sparkly speed-line background, speech bubble in Japanese: 「みんなー！マンガやイラストをもっと自由に楽しみたい？それなら…」, name label: 「アイくん」. Panel 2, top-right logo panel: large playful purple and blue “AniFusion” logo with a painter palette incorporated into the letters, Japanese subtitle: 「アニフュージョン」. Panel 3, upper-right small mascot panel: Ai-kun waves beside the logo with speech bubble: 「AniFusionがすべて叶えてくれるよ！」. Panel 4, middle large dashboard panel: central web-app screenshot mockup labeled “Welcome back”, surrounded by Ai-kun presenting it; top jagged caption: 「いろんなツールがぜ〜んぶここに！」. Panel 5, left stack tool card area: exactly 3 stacked feature cards labeled “Comics”, “Canvases”, and “Image”; each has thumbnail grids and Japanese captions respectively: 「マンガを作ろう！」, 「アイデアを広げよう！」, 「高品質なAI生成！」. Panel 6, lower-middle feature card: “Video” red label with a dynamic anime boy thumbnail and Japanese caption: 「イラストを動画に！」. Panel 7, lower-middle feature card: “Drawing” orange label with sketch-to-finished-anime-girl artwork and Japanese caption: 「描いて、整えて、仕上げよう！」. Panel 8, lower-right feature card: “Library & More” green label with exactly 4 simple black icons: bookshelf, open box/cube, checklist/document, sliders/settings; Japanese caption: 「保存も管理もカンタン！」. Panel 9, bottom two-part scene: left side Ai-kun hugs a Shiba Inu dog on a pink heart background with speech bubble: 「AIがキミの創造力をサポートするよ！」; right side Ai-kun sits at a desk with drawing tablet, stylus, monitor showing manga page thumbnails, coffee mug with paw print, and speech bubble: 「さあ、いっしょに楽しい作品を生み出そう！」. Right vertical strip: Add a narrow white vertical strip on the far right with black Japanese text reading: 「これはよくある案件漫画のフリをしたコンテスト応募作品です(ChatGPT free)」 followed by the name 「大澤基」 and a small blue paw print. Bottom banner: Full-width light blue banner with paw prints and stars, large orange-and-white Japanese headline: 「クリエイターのための、ぜんぶ入りプラットフォーム！」 and smaller line: 「AniFusionで、きみの物語をカタチにしよう！」. Visual style: Modern polished anime manga advertisement, glossy cel shading, expressive faces, clean line art, halftone textures, sparkles, starbursts, pastel blue and peach backgrounds, readable Japanese typography, friendly creator-tool branding. Constraints: Include exactly 6 visible appearances of Ai-kun: pointing close-up, waving by logo, presenting dashboard, small pointing mascot near feature cards, hugging dog, and drawing at desk. Include exactly 3 stacked left feature cards, exactly 3 bottom-row feature cards, and exactly 4 library icons. Keep all Japanese text legible and placed as described. No watermark, no extra characters beyond Ai-kun and the dog.

生成一组四格日式温情漫画，干净手绘动漫线稿风格，温暖米棕色与浅棕色单色调，柔和奶油色背景，简单柔和阴影，整体温馨治愈氛围。画面比例：正方形1:1，标准2x2四格布局，每格之间用细棕色边框分隔。顶部标题：大字「{argument name="标题" default="夫婦の日"}」，左右各加一颗小心心，手写感深棕色。 【四格内容】 第1格：年轻亚洲情侣站立肖像。{argument name="男性角色" default="男人"}站在{argument name="女性角色" default="女人"}身后，双手轻轻放在她肩上，两人微笑看向镜头。 第2格：情侣坐在温馨室内，双手捧着马克杯，相视而笑。女人对话框清晰写着中文：「谢谢你一直陪着我。」白色圆角对话框，黑色清晰文字。 第3格：情侣一起在厨房做饭。两人并肩站在灶台前，男人搅拌，女人帮忙，表情专注又幸福。 第4格：情侣紧紧拥抱。男人低头轻轻亲吻女人的额头，女人闭眼微笑。 【角色一致性强制要求】 两个角色从第1格到第4格必须保持完全同一张脸、同一发型、同一服装、同一身材比例。禁止任何漂移。 男人：短发、温和长相、穿简单毛衣或衬衫。 女人：中长发、温柔长相、穿家居感衣服。 【风格与质感】 干净手绘动漫线稿，单色调温暖米棕+浅棕+奶油白，简单阴影，像日式生活漫画或治愈系插画。 【文字要求】 对话框内中文必须正确、清晰、无乱码、无变形。避免：写实摄影、3D渲染、鲜艳色彩、复杂背景、角色脸崩、文字乱码、廉价贴纸感。

Two-dimensional alternative character illustration of {argument name="character" default="the subject"}, occupying approximately seventy-five percent of the frame in {argument name="pose" default="a dynamic stance"}, with [HAIR] and [EXPRESSION]. Oversized rounded cranium, a short flattened lower face, a narrow jaw, a long thin neck, slightly oversized ears, huge pale oval eyes with tiny off-center pupils, extremely thick horizontal black eyebrows, heavy black upper eyelids, muted mauve under-eye rings, a tiny broad triangular nose, and a wide asymmetrical grin showing individually drawn irregular rectangular teeth, one clearly missing upper tooth, one chipped tooth and deep black gaps. Dress the character in {argument name="clothing" default="simplified graphic clothing"}, simplified into broad graphic shapes with chunky seams, minimal folds and small hand-drawn accessories. Confident black ink contours using thick outer silhouettes, medium-weight facial lines, thin angular hair strands and only a few short scratch marks around the eyes, hair and clothing seams; use flat opaque color fields, one hard-edged cel-shadow shape per form, sparse pencil hatching, subtle ink wobble, slight color misregistration and restrained worn-paper texture. Controlled [COLORS] palette, keeping every color clearly separated and moderately saturated. Place the character against [BACKGROUND], reduced to large flat shapes, a few loosely drawn environmental details and oversized rough hand-painted lettering reading “[TEXT]”; use an off-center waist-up composition, a slight camera tilt, cropped background typography, loose poster margins and lightly weathered printed edges. 4:5 AR.

Goal: Create a cinematic dark-fantasy character poster for {argument name="character name" default="Victor Von Doom"}, styled like a premium comic-book armor reveal and collectible cover. Canvas: Horizontal 16:9 composition, gritty off-white paper background on the left and a large close-up armored portrait dominating the right two-thirds. Use high contrast, realistic rendering, dramatic shadows, halftone print grain, subtle scratches, and distressed poster texture. Main subject: A hooded sovereign villain in ornate black forged-latverian armor, shown from the shoulders up in an intense three-quarter close-up. He wears a deep {argument name="hood color" default="emerald green"} woven fabric hood that casts heavy shadow over his face. His metal mask is dark gunmetal with engraved filigree, rivets, sharp angular plates, and a heavy central nose bridge. Add an elaborate spiked spectacle-like eye frame across the mask, with two reflective lenses: the left lens glowing green and the right lens glowing gold. One human eye is barely visible through shadow above the lenses, sharp and intimidating. Shoulder armor is glossy black with gold ornamental trim and a crest-like filigree emblem. Include a dark green leather-gloved hand or armored forearm partially visible in the lower right foreground. Layout: The left third is a graphic poster margin with vertically oriented typography. Place one broad diagonal paint stroke in dark green behind the portrait, sweeping from upper center-left toward lower right. Include exactly 1 small rectangular inset image near the lower center-left showing a cropped close-up of the spiked green-and-gold lenses, framed by a thin gold border. Text content: Include exactly 5 visible text elements: 1) vertical tall condensed black title reading "{argument name="vertical title text" default="GENTLE MONSTER × DOCTOR DOOM"}" along the far left; 2) large bold model label reading "{argument name="model label" default="DOOMWAR VV-616"}" below it; 3) three-line small spaced uppercase tagline reading "ENGINEERED FOR DOMINION. / FORGED IN LATVERIA. / WORN BY SOVEREIGNS."; 4) tiny gold signature in the lower right reading "{argument name="signature text" default="Victor Von Doom"}"; 5) a small gold ornamental crest centered in the left margin between the vertical title and model label. Visual style: Ultra-detailed photorealistic dark fantasy mixed with luxury fashion advertising, Marvel-inspired villain aesthetics, antique metalwork, green and gold color accents, brutalist typography, premium editorial poster design, dramatic studio lighting, shallow depth of field, razor-sharp mask details. Constraints: Keep the composition asymmetrical with the face cropped close on the right edge, preserve the exact count of 1 inset image and 5 text elements, use no extra logos or watermarks, and avoid cartoon or anime styling.

Create a hyper-realistic, cinematic editorial portrait of the {argument name="subject" default="same young woman with long wavy dark-brown hair, soft natural makeup, and a calm, confident expression"}. She wears a {argument name="clothing" default="blue denim jacket over a fitted top, matching blue jeans, and black high heels"}, maintaining the exact same facial identity, hairstyle, body proportions, and outfit throughout the composition. Blend two surreal concepts into one seamless scene: On one side, depict the woman as a giant standing waist-deep in a modern international shipping port, surrounded by container terminals, cargo cranes, warehouses, turquoise seawater, and multiple cargo ships. She gently holds a miniature container ship in one hand while several real cargo vessels navigate around her, leaving realistic water wakes. A dramatic overcast sky with dark storm clouds creates a cinematic atmosphere, with a futuristic city skyline visible in the distance. On the other side, merge the environment into a black-and-white Japanese manga collage wall filled with dynamic comic panels featuring motorcycles, office scenes, city streets, livestreaming, friendship, expressive characters, speed lines, Japanese sound effects, and speech bubbles. The woman stands naturally in front of the manga artwork with a relaxed editorial pose, creating a striking contrast between the colorful realistic subject and the monochrome manga background. The transition between the shipping port and manga collage should feel smooth and artistic, combining surreal scale, fantasy, editorial fashion photography, and comic-book aesthetics into one cohesive composition. Ultra-detailed textures, realistic lighting, sharp focus, natural skin tones, HDR, 8K resolution, volumetric lighting, cinematic color grading, professional fashion photography, highly detailed environment, masterpiece quality.

Goal: Create a black-and-white manga page about a late-night annoying interruption, with the title {argument name="manga title" default="深夜のウザ絡み"} at the top left and page indicator {argument name="page indicator" default="1/3"} at the top right. Canvas: Vertical manga page, monochrome ink and screentone, clean white gutters, detailed crosshatching, elegant shoujo/seinen line art. The page contains exactly 5 numbered panels, each with a small boxed panel number in the corner. Main character: A beautiful young male calligrapher, pale skin, delicate tired eyes, long light-colored wavy hair with a thick braid over one shoulder, wearing layered traditional Japanese robes/kimono. He sits at a low wooden writing desk with inkstone, brush, papers, stacks of books, shelves of old books, and a quiet Japanese study atmosphere. Second character: A cute round hamster, fluffy and expressive, acting as an intrusive announcer. Use {argument name="hamster announcement" default="ハムスターによる、ウザ絡みー"} as the large dramatic announcement text in the hamster panel. Layout and panel count: 1. Wide top establishing panel across the page: night scene inside a traditional Japanese room. Sliding doors open to a snowy garden under a full moon; snow-covered trees and stone lantern outside. The calligrapher writes at his desk, surrounded by books and wall scrolls. Add small handwritten sound text near his brush: {argument name="writing sound effect" default="さらさら…"}. 2. Middle-right large comedic panel: the hamster stands on the manuscript with arms spread, sparkling background and dramatic radial burst. Add a vertical speech bubble reading {argument name="opening phrase" default="深夜のー"}, and the large announcement text beside it. 3. Middle-left narrow panel: close view of the calligrapher looking down at his paper, silent and unimpressed. Add a small speech bubble with only ellipsis: "……". 4. Bottom-right panel: the calligrapher rests his cheek on one hand, bored and mildly annoyed at the desk. Add a speech bubble with Japanese text meaning he is not good at literary teasing but it is fine: "文机の上は舞台じゃないんだけどな。" Add a small sigh sound effect: "ふぅ…". 5. Bottom-left panel: close dramatic panel of the calligrapher looking down at the hamster, who is seen from behind at the bottom edge. Use a speech bubble from the calligrapher: {argument name="final dialogue" default="……で、こんな時間にわざわざボーズまで決めて。何かあった？"}. Visual style: Highly detailed Japanese manga, grayscale only, fine pen strokes, realistic textures for wood, paper, snow, hair, fabric folds, books, and garden. Balance serious late-night historical atmosphere with absurd cute-hamster comedy. Keep all text hand-lettered in Japanese manga style, preserve the exact 5-panel structure, no color, no watermark, no extra panels.

Create a cinematic vertical street photograph in a gritty urban alley at golden hour, showing a real young goth-punk woman with {argument name="hair color" default="black teased shaggy hair"} standing on the left side of a concrete utility pole and leaning toward a wall mural on the right as if flirting with or meeting her illustrated alter ego. She wears a dark oversized plaid flannel shirt, black layered top, studded belt, ripped black skinny jeans, fingerless gloves, layered necklaces, heavy eyeliner, dark lipstick, and tall black platform lace-up boots with buckles; one boot is lifted behind her in a playful pose, her hands resting on the pole, gaze turned toward the drawing. The environment is a narrow weathered alley with cracked asphalt, rough stucco walls, metal pipes, stickers, faded graffiti tags, grime, and warm backlit sunlight creating long shadows and shallow depth of field. On the right wall, draw exactly 1 large black ink graffiti-style cartoon character matching her punk aesthetic: wild long black hair, huge laughing smile, plaid shirt, ripped pants, platform boots, one leg kicked up, leaning toward the pole; surround it with exactly 9 doodle elements: 2 hearts above, 1 skull speech bubble, 3 five-point stars, and 3 short motion/emphasis line clusters. Make the mural look hand-drawn directly on the wall in messy expressive black marker lines, while the real person remains photorealistic. Use high-detail textures, natural lens blur, moody fashion-editorial lighting, rebellious cute punk energy, and a realistic color grade with muted browns, charcoal blacks, and warm highlights. No extra people, no readable brand logos, no watermark.

Using REFERENCE_0 as the character base, expand it into a full illustrated outdoor scene while keeping the same bold black outline, hand-painted texture, and recognizable character design. Change the character from a static standing pose into a walking pose, holding exactly 1 red leash in their right hand, with exactly 1 sweat drop on the face and a slightly tired expression. Add exactly 1 small dog, {argument name="dog breed" default="Shiba Inu"}, walking in front of the character on the leash, happy and energetic with its mouth open and curled tail. Place them in {argument name="setting" default="a sunny riverside path"}: blue river behind them, green trees along the banks, sandy ground with scattered dark stones, and a bright blue sky. Add exactly 1 white cartoon cloud in the upper left. Add exactly 1 white thought bubble in the upper right with exactly 2 small connector dots, containing vertical Japanese text {argument name="thought bubble text" default="あつい..."}. Keep the composition landscape-oriented, with the dog on the left and the character on the right, both facing left as if walking together. Maintain a chunky oil-paint/impasto surface texture, vivid colors, simple cartoon shapes, thick black borders, and no photorealism.

Create a close-up meme-style digital illustration of {argument name="character" default="Pepe the Frog"} from the shoulders up, filling most of the frame. The character has a large green frog head with smooth rounded cheeks, a wide flat face, and a small calm smirk with pale pink lips. He wears oversized oval glasses with thin gray rims; both lenses are extremely bright white with a soft bloom effect, obscuring the eyes as if glowing with intense reflection. One green hand is raised on the right side of the image, with the index finger touching or pushing up the bridge of the glasses in a clever, smug gesture. The character wears a plain {argument name="shirt color" default="bright blue"} shirt visible at the bottom. Use thick uneven black sketch outlines, simple flat cartoon coloring, soft airbrushed green shading on the face and hand, and a low-detail gray {argument name="background color" default="light gray"} background. Composition should be tightly cropped, slightly off-center, with the glasses glow as the brightest focal point, matching an internet meme / reaction image aesthetic.

Goal: Create a cinematic collectible character poster for a dark fantasy / superhero crossover villain, featuring {argument name="character name" default="Gentle Monster × Doctor Doom"} as the main vertical side title and {argument name="model name" default="DOOMWAR VV-616"} as the product-style designation. Canvas: Wide 16:9 landscape poster, gritty off-white parchment background with halftone speckling, black ink wear, and a muted yellow-green tint. The composition is split: a narrow typography column on the left and a huge close-up character portrait filling the center and right. Main subject: One armored hooded ruler shown in extreme close-up from shoulders up, turned slightly toward camera. He wears a deep forest-green fabric hood with coarse woven texture, a dark gunmetal medieval sci-fi mask, ornate rivets, engraved filigree, sharp cheek and brow plates, and a sculpted metal nose bridge. His eyes are partly visible in shadow behind the mask, intense and human. Across the eye area, add one elaborate spiked eyeglass frame integrated into the mask, with exactly two mirrored lenses: left lens emerald green, right lens golden yellow. The eyewear is jagged, thorn-like, chrome-silver, with high specular highlights. The armor includes blackened steel plates, gold ornamental trim, green leather gloves or shoulder material at the lower right, and one visible gold crest emblem on the chest/shoulder armor. Layout and countable elements: Include exactly 1 main portrait, exactly 1 small rectangular inset image near the lower center-left showing a close-up of the same spiked green-and-gold glasses, exactly 1 vertical title stack on the far left, exactly 1 product designation block below it, exactly 1 short technical description block under the designation, exactly 1 thin vertical gold rule along the far left, exactly 1 small gold ornamental crest below the rule, and exactly 1 tiny signature in the bottom-right corner reading {argument name="artist signature" default="Victor Von Doom"}. Text content: On the far left, set the vertical black condensed uppercase text to read {argument name="vertical title text" default="GENTLE MONSTER × DOCTOR DOOM"}. Beneath it, use a bold horizontal product label reading {argument name="designation text" default="DOOMWAR VV-616"}. Under that, place three short uppercase lines: “ENGINEERED FOR DOMINION.”, “FORGED IN LATVERIA.”, “WORN BY SOVEREIGNS.” Use tall condensed sans-serif typography, distressed print texture, and dark charcoal ink. Visual style: Hyper-detailed cinematic realism, luxury eyewear campaign meets comic-book villain armor, macro photography lighting, dramatic shadows under the hood, subtle film grain, halftone poster texture, desaturated green and gold palette, metallic reflections, premium editorial layout, sharp focus on mask and lenses, background brush stroke of dark green paint behind the head. Constraints: Keep the face largely hidden by the hood and mask, do not add extra characters, do not add additional inset images, keep all visible text legible and limited to the specified text, no watermark.

Goal: Create a vertical 2-column by 3-row collage of exactly 6 separate illustration panels, all in a unified retro Japanese modernist poster style with flat geometric shapes, hard-edged composition, limited palettes, subtle paper grain, ink texture, and strong black/cream/red/orange accents. Each panel should look like a distinct generated artwork sharing the same visual language. Canvas: Tall portrait canvas, approximately 675 × 1024 px or 2:3 ratio. Use thin white gutters between panels and a narrow white outer border. No captions, no logos, no watermark. Visual style: A fusion of mid-century graphic design, Japanese printmaking, Constructivist geometry, art deco poster composition, and contemporary anime/editorial illustration. Use flattened perspective, blocky shapes, textured paper, muted cream backgrounds, deep black shadows, burnt orange, vermilion red, slate gray, charcoal, and occasional navy blue. Keep details stylized rather than realistic. Layout: Arrange exactly 6 panels: 1. Top-left panel: A geometric night scene with an orange kangaroo-like animal standing upright beside a gray koala-like animal, a cream full moon in the upper left, small star/firework dots, and a grid of orange, beige, black, and gray rectangles in the background. The animals should be angular and origami-like. 2. Top-right panel: A dark gothic anime-style girl with long pale hair sitting on stone steps before ruined cathedral arches. Use black, gray, lavender, and hot magenta accents, with one red eye visible and a moody ruined-city atmosphere. 3. Middle-left panel: A reclining elegant long-haired person in black clothing holding a round fan near their face, surrounded by red and black geometric circles, grids, rectangles, and cream negative space. The mood is languid and stylish. 4. Middle-right panel: A close-up portrait of a pale bob-haired girl with straight black bangs, red eyes, small red vertical face markings, and a black-white-red abstract geometric background with circles and rectangular blocks. 5. Bottom-left panel: A minimalist city-walking scene with a person in a long gray coat holding a leash attached to a low, stretching orange-and-white corgi. Background includes red circles, black blocks, cream walls, stairs, and sharp architectural geometry. 6. Bottom-right panel: A back-view portrait of a traditional Japanese woman or geisha in a patterned kimono, standing before pagoda rooftops and a lantern. Include a bright blue sky with cream clouds and bold black, red, cream, and tan architectural shapes. Subject customization: Keep the collage structure fixed, but allow the overall subject set to be interpreted as {argument name="theme" default="six unrelated retro Japanese poster scenes"}. Use {argument name="main palette" default="cream, black, vermilion red, burnt orange, charcoal gray, navy blue"}. Set the mood to {argument name="mood" default="stylish, mysterious, nostalgic, graphic"}. Use {argument name="art style" default="retro Japanese modernist poster illustration with anime influence"}. Make the image suitable for {argument name="use case" default="a social media showcase of varied GPT Image 2 outputs from one consistent style prompt"}. Constraints: Exactly 6 panels, exactly 2 columns and 3 rows, no text inside the artwork, no photorealism, no extra panels, no floating UI, no frame labels. Maintain consistent texture and design language across all six panels while making every panel visually distinct.

Goal: Create a vertical fantasy trading card for {argument name="character name" default="LANDCASTER"}, styled like a premium dark collectible RPG card titled “My Fantasy Trading Card,” with gothic cyberpunk necromancer energy. Canvas: Portrait trading card, 2:3 ratio, highly detailed, full bleed with an ornate blackened-gold frame. Use a dark palette of black, midnight blue, violet, magenta neon, and antique gold, with glowing magical effects and sharp high-contrast fantasy illustration rendering. Layout: At the top, place a large glowing serif nameplate reading {argument name="character name" default="LANDCASTER"}. Beneath it, a smaller subtitle banner reading {argument name="class subtitle" default="NEON GRAVEBINDER"}. Add a circular skull faction icon at the top left and a round level badge at the top right reading {argument name="level" default="VII"}. The center contains one main character portrait. The lower third contains a rarity banner, class label, six stat boxes, two ability text boxes, a quote box, and bottom tags. Subject details: The main portrait is a dark-skinned young adult male shadow mage with shoulder-length dreadlocks, wearing a purple knit beanie, a dark blue leather jacket with metal studs and pins, and a white skull-pattern face mask covering the lower half of his face. His eyes glow bright violet-white with supernatural light. He faces forward in a confident, intimidating pose. Add small skull accessories, occult ornaments, and purple magical highlights around him. Background: A neon graveyard city at night, with blue and purple skyscrapers, glowing Japanese-style vertical signs, arcane rune circles behind the character, spectral energy, floating particles, and skull motifs. The mood is urban fantasy, cursed, electric, and mystical. Card frame and ornaments: Use an elaborate gothic border with gold filigree, black lacquer panels, thorn-like metal flourishes, small skulls at corners and dividers, purple crystal accents, and a central skull gem medallion above the rarity banner. Make the card look like a high-value foil holographic collectible. Visible text content: Top name “LANDCASTER”; subtitle “NEON GRAVEBINDER”; level “VII”; rarity banner “EPIC”; class label “SHADOW MAGE”. Include exactly 6 stat boxes: HP 2780 with a heart icon, ATK 342 with a sword icon, DEF 251 with a shield icon, MAG 389 with a purple flame icon, AGI 364 with a wing icon, LCK 207 with a clover icon. Include exactly 2 ability entries: “SOUL CHAIN — Bind enemies in spectral chains, dealing MAG damage over time and reducing their DEF.” and “GRAVEBOUND AURA — Increases MAG and AGI. Restores HP when defeating a cursed enemy.” Add a quote box reading {argument name="quote" default="The dead don’t speak to me. I speak for them."}. Bottom tags: “CURSED” on the left and “NIGHT CITY ORIGIN” on the right. Style constraints: Ultra-detailed fantasy card art, crisp readable typography, luminous purple glow, dramatic rim lighting, painterly realism mixed with comic/anime trading-card polish, no extra characters, no watermark, no modern UI outside the card, keep all listed text legible and arranged like a real collectible game card.

Use the uploaded reference image as the strict identity and outfit reference for a realistic {argument name="subject" default="Japanese woman"}, preserving her face identity, facial proportions, eye shape, nose, lips, skin tone, hairstyle, hair colour, visible accessories, overall vibe, and outfit exactly as shown in the reference. Do not invent or hardcode features that are not present in the uploaded image. Create a high-quality vertical mixed-media portrait featuring a realistic full-body Japanese woman standing beside a black {argument name="shadow style" default="hand-drawn doodle-shadow"} version of herself sketched directly onto a {argument name="background" default="clean wall"}. The real woman should remain photorealistic, wearing the same outfit and styling as the reference, with a cute, slightly confused, mildly embarrassed expression as if wondering why she is playing along with her mischievous shadow. The doodle-shadow should be a lively black sketch version of the same Japanese woman, matching her hairstyle, outfit silhouette, and overall identity while having exaggerated cartoon expressions and energetic movements, surrounded by manga-style motion lines, hearts, sparkles, stars, and comic symbols. For every generation, invent a completely new playful pose where both the real woman and the doodle share the same general idea without perfectly mirroring each other. Avoid repeated pointing poses, finger-guns, or stiff standing poses, and instead create fresh playful actions such as dancing, balancing, sneaking, making a giant heart, peekaboo, dramatic overreaction, or other cute troublemaker poses. Place the real woman on one side of the frame and the doodle-shadow on the wall beside her so their interaction is immediately clear. Use a clean white, cream, or light grey indoor wall with minimal clutter, soft natural daylight, gentle shadows, and a polished editorial photography style. The final image should be ultra-detailed, social-media-friendly, with strong resemblance to the uploaded Japanese woman, faithful outfit reproduction, realistic photography combined with expressive black sketch line art, premium composition, and no AI artifacts. Avoid unrelated outfit changes, realistic second people, ordinary shadows, horror elements, coloured illustrations, anime humans instead of doodles, weak resemblance, cluttered backgrounds, repeated poses, text, logos, watermarks, distorted anatomy, extra limbs, extra fingers, or poorly drawn hands.

Goal: Using REFERENCE_0 as the manga name/storyboard, convert it into a polished finished manga page while preserving the same page structure, panel order, speech-balloon placement, and story beats. Canvas and layout: Create a clean vertical black-and-white manga page with full-color illustration. Keep exactly 4 panels: 1 wide top panel, 2 diagonal split middle panels, and 1 wide bottom panel. Add the header title {argument name="title text" default="夜食"} at top left, page number {argument name="page number" default="2/4"} at top right, and artist handle {argument name="artist handle" default="@_sagyo"} centered at the bottom. Transformations from the reference: Replace the rough placeholders with finished anime-style character art. The top panel should show the ordinary girl holding a bowl of tofu with sauce in a warm kitchen/dining setting. The middle-left placeholder labeled as the demon version should become a beautiful silver-haired demon man with black horns, dark wings, irritated expression, and a purple ominous background. The middle-right placeholder labeled as the angel version should become the same silver-haired man as an angel, smiling with clasped hands, white wings, halo, flowers, sparkles, and a soft golden background. The bottom panel should show the girl holding a small shichimi spice bottle, with the remaining dialogue effects and balloons arranged like the storyboard. Text: Use exactly 9 visible speech/effect balloons total, matching the reference positions: top panel has 1 balloon with {argument name="top panel dialogue" default="じゃ…\nお豆腐に\nめんつゆかけて…！"}; middle-left panel has 2 balloons with {argument name="demon panel dialogue" default="ちっ…\n冷奴は正直おいしいので\n何も言えません… / せめてネギとおかかは\n載せましょうよ\nそれくらいの悪は\n許されます"}; middle-right panel has 2 balloons with {argument name="angel panel dialogue" default="まぁ…！\n夜食界の\n優等生ですよ… / えらいです…"}; bottom panel has 4 balloons/effects with {argument name="bottom panel dialogue" default="七味も…\nかける…！ / くっ…\n今夜は天使側の\n勝ちですか… / それは悪ではなく\n彩りです / 七味"}. Style: High-quality polished Japanese web-manga rendering, clean lineart, soft gradients, expressive faces, accurate balloon lettering, crisp black panel borders, no rough storyboard guides, no blue placeholder boxes, no panel-number markers, and no extra characters or extra text.

Two-dimensional alternative character illustration of {argument name="character" default="[CHARACTER]"}, occupying approximately seventy-five percent of the frame in {argument name="pose" default="[POSE]"}, with {argument name="hair" default="[HAIR]"} and [EXPRESSION]. Oversized rounded cranium, a short flattened lower face, a narrow jaw, a long thin neck, slightly oversized ears, huge pale oval eyes with tiny off-center pupils, extremely thick horizontal black eyebrows, heavy black upper eyelids, muted mauve under-eye rings, a tiny broad triangular nose, and a wide asymmetrical grin showing individually drawn irregular rectangular teeth, one clearly missing upper tooth, one chipped tooth and deep black gaps. Dress the character in {argument name="clothing" default="[CLOTHING]"}, simplified into broad graphic shapes with chunky seams, minimal folds and small hand-drawn accessories. Confident black ink contours using thick outer silhouettes, medium-weight facial lines, thin angular hair strands and only a few short scratch marks around the eyes, hair and clothing seams; use flat opaque color fields, one hard-edged cel-shadow shape per form, sparse pencil hatching, subtle ink wobble, slight color misregistration and restrained worn-paper texture. Controlled [COLORS] palette, keeping every color clearly separated and moderately saturated. Place the character against [BACKGROUND], reduced to large flat shapes, a few loosely drawn environmental details and oversized rough hand-painted lettering reading “[TEXT]”; use an off-center waist-up composition, a slight camera tilt, cropped background typography, loose poster margins and lightly weathered printed edges. 4:5 AR.

Create a direct contrast between traditional pencil sketches and high-purity digital lighting effects. * Aggressive graphite cross-hatching, smudged charcoal Skin, fabric, and armor are all shaped using visible cross-hatching, graphite smudges, and charcoal smudges; the rough texture of the paper is preserved, and no smooth shading is applied. * Pure white negative space, cool-gray monochrome base Keep the background clean and pure white, with the subject rendered in cool gray and blue-gray tones; the more white space left, the more the line art and modeling will resemble a high-end RPG art book. * blinding flat amber glow, unblended digital overlay Allow the amber glow to appear only suddenly in the form of cracks, particles, and light sources, with hard edges and pure color; avoid soft gradients—the clash between cool gray and golden light is what makes the composition truly impactful. Create a completely original ultra-premium character concept art illustration with a newly invented theme: "The Amber-Bound Eidolon," a powerful, athletic female phantom guardian holding a vintage glass astrolabe filled with glowing golden moths, with her lower body dissolving into smudged charcoal and glowing golden sand. The subject and lore must be fully original, but the visual style must maintain 100% exact fidelity to high-end RPG tabletop character design sheets, isolated on a pure white background. The absolute highest priority is the extreme visual collision between raw, dry, cool-grey pencil sketching (graphite hatching) and blinding digital luminescence. No environment, no readable text, no logo. Style direction: This is the absolute priority. Fully recreate the exact image-family style: premium fantasy character concept art isolated on a stark white negative-space background, heavily emphasizing "Graphite Hatching" and "Smudged Charcoal" effects. All shadow areas on the female character's skin, armor, and clothing must feature highly visible, aggressive pencil cross-hatching. The illustration must feel like a master-level traditional graphite drawing on textured paper that has been violently interrupted by blazing, flat digital colors. The contrast between the extremely dry, cool-grey pencil rendering and the piercing, neon-bright digital amber glow is the core of the image. Core composition: Use a tall vertical composition focused entirely on a single female character floating in a void of pure white paper space. The figure is posed dynamically: one arm is raised, holding a heavy, dangling, brilliantly glowing glass astrolabe. Her upper torso is muscular and solidly drawn, while from the waist down, her draped clothing and legs completely unravel and disintegrate into rising, stylized smoke. This smoke must look like aggressive charcoal smudging mixed with glowing, particle-like golden sand fading into the white background. Theme & Character design: Design the Amber-Bound Eidolon: She must have a strong, healthy Western athletic female physique with firm, defined muscle lines (strong shoulders, sculpted core), projecting fierce dignity and power—absolutely avoiding any generic, soft, or "internet-celebrity" facial features. Her body is rendered almost entirely in cool slate-grey and ash-blue pencil tones. Deep, piercing, blindingly bright cracks of liquid amber run across her athletic chest, shoulders, and arms. She wears tattered waist wrappings and extremely intricate, aged bronze bracers featuring razor-sharp pencil filigree. Prop design: Hanging from her hand is a massive, ornate vintage glass astrolabe or lantern. Its interior blazes with a blinding, intensely hot neon-gold and pure yellow light, swarming with magical glowing moths. Wisps of golden heat drift upward from it, casting a harsh, luminous orange-gold rim-light that cuts aggressively across the dry graphite textures of her arm and athletic torso. Texture and finish: The surface tactility is everything. The white background must be pure and clean. The character's matte surfaces must show visible paper grain, dry graphite drag, and raw hatching lines. The glowing liquid amber cracks and the astrolabe fire must look like pure, unblended, intensely bright digital overlays that pop aggressively against the desaturated sketch. Mood: Epic, raw, aggressive contrast, dark fantasy, powerful feminine strength, imposing, highly detailed studio-grade sketch. Rendering style: ultra-premium RPG character concept art, isolated on pure white background, extreme graphite cross-hatching, smudged charcoal smoke, blinding digital amber glow, stark cool-grey vs neon-gold contrast, intricate bronze filigree, athletic female warrior, lower body dissolving into pencil strokes and golden sand, world-class illustration, 8k. Negative prompt: background environment, scenery, readable text, logo, photorealism, 3D render, smooth airbrushing, plastic texture, generic beauty face, weak anatomy, soft anime girl, full solid legs, flat cartoon shading, muddy colors, soft glowing gradients.

Goal: Create a dynamic comic-book storyboard/collage showing an anime-style futuristic superhero battle on a rainy cyberpunk rooftop at night, featuring {argument name="spider hero name" default="a black-and-red spider-themed hero"} and {argument name="armored hero name" default="a blue armored tech hero with a round shield"} fighting surveillance drones and laser attacks. Canvas: Square 1:1 image, arranged as a polished comic page with thick clean white gutters separating exactly 7 panels. Use cinematic lighting, high-contrast neon reflections, rain streaks, wet rooftop puddles, storm clouds, and a neon city skyline in the background. Overall mood: intense, heroic, dramatic, action-heavy. Layout: Use exactly seven discrete panels: 3 narrow panels across the top row, 1 large wide action panel across the middle row, and 3 smaller panels across the bottom row. Panel details: 1. Top-left panel: wide rooftop establishing shot. The spider-themed hero crouches low in the foreground on a wet rooftop, black suit with red web patterns and glowing white eye lenses; the armored hero stands farther back. Neon signs, antenna towers, rain, and glossy reflections fill the scene. 2. Top-middle panel: close-up portrait of the spider hero’s masked face under a red hood, white glowing eyes, red web pattern, clenched red fist raised beside the face, red lightning-like web energy crackling around the hand. 3. Top-right panel: close-up portrait of the armored hero, sleek dark blue and metallic armor, cyan glowing visor, circular cyan chest reactor, raised round shield with red-white-blue rings, heroic three-quarter view. 4. Middle large panel: main battle scene. On the rooftop, the spider hero slides or lunges through splashing rainwater while projecting a glowing cyan circular energy web-shield from one hand toward the armored hero. The armored hero stands braced with cape flowing, shield ready, cyan energy beam firing upward. Multiple drones hover in storm clouds, shooting red laser beams across the panel; sparks, water splashes, and neon city buildings intensify the action. 5. Bottom-left panel: the spider hero crouches alone on the wet rooftop, pointing one hand forward as if preparing to shoot a web; rain, antenna mast, and dark buildings behind. 6. Bottom-middle panel: the armored hero stands in a battle stance holding the round shield while generating a bright cyan energy sphere; concentric red and cyan circular energy rings swirl around him. 7. Bottom-right panel: both heroes pose together on the rooftop beneath a giant glowing red spiderweb spread across the sky, with a blue drone caught at the center; the spider hero crouches in front while the armored hero stands behind holding the shield. Visual style: Modern anime superhero illustration mixed with American comic-book paneling, sharp clean line art, saturated neon cyan and red effects, glossy wet surfaces, dramatic rim lighting, detailed armor highlights, expressive action poses, cinematic depth, rain-soaked cyberpunk atmosphere. Text content: No speech bubbles, captions, logos, or readable text. Constraints: Keep the exact 7-panel comic layout and the two main heroes consistent across all panels. Use {argument name="primary energy color" default="cyan"} for the armored hero’s technology effects and {argument name="secondary energy color" default="red"} for lasers and spider energy. Background should remain a rainy neon rooftop city at night throughout.

Goal: Create a black-and-white shojo horror manga page titled {argument name="manga title" default="籠目"}, page indicator {argument name="page number" default="1/8"}, adapting an eerie scene in an old Japanese mansion corridor. Use delicate manga linework, soft gray screentones, high-detail wood grain, romantic character rendering, and unsettling quiet horror atmosphere. Canvas: Vertical manga page, portrait orientation, white outer margin, thin black page border, with the title at the top left and page number at the top right. Monochrome only, printed manga texture, no color. Layout: Use exactly 4 panels separated by thick black gutters. Panel 1 occupies the full width across the top half of the page. Panel 2 is a full-width narrow horizontal panel underneath. Panel 3 is the lower-left large portrait panel. Panel 4 is the lower-right tall corridor panel. Panel 1: A deep one-point-perspective hallway in an old Japanese house, with shoji windows along the left side casting rectangular lattice shadows across the wooden floor. At the far end are closed sliding doors. Centered in the hallway, seen from behind, stands a slender young person in a kimono and hakama, with long pale hair tied into a braid down the back. The figure is still, facing the closed doors. Add exactly 2 vertical narration boxes: one at the upper right and one at the lower left. The upper-right box says in Japanese vertical text: "奥の間の襖を引くとき、私はいつも少しだけ息を止める". The lower-left box says: "開けた瞬間の空気の温度を、確かめる癖がついてしまった。" Panel 2: Extreme horizontal close-up from the side of the same androgynous beautiful character’s face and hand. Their damp-looking light hair falls over their eyes, expression calm but tense, wearing an earring. One hand reaches toward or touches a round sliding-door pull. Add exactly 1 vertical narration box on the right saying: "三月ほどかけて整えた。誰にも言わずに。" Panel 3: Lower-left close portrait of the same character from the chest up, elegant shojo style, long pale wavy hair gathered into a side braid, one hand thoughtfully touching the chin. The background is very dark with a subtle cage-like or hexagonal wire pattern, adding horror symbolism. Add exactly 1 white vertical text box on the left with large Japanese text: "そう" and smaller text below: "造った。" Panel 4: Lower-right panel showing an empty old wooden corridor and sliding wall in perspective, with shoji light spilling into darkness; the far hallway fades into black. Include exactly 2 vertical narration boxes: one near the upper right saying: "駐屋の書斎と地続きでありながら、" and one lower left saying: "この部屋にだけ隠れていた本余計に折れている。" Subject details: The character is a youthful, pale, delicate shojo-manga protagonist with {argument name="hair color" default="silver-white"} messy long hair, a thick braid, fine eyelashes, small earrings, and a dark traditional Japanese outfit. Their mood is quiet, secretive, and haunted rather than overtly frightened. Visual style: Highly detailed monochrome manga, fine pen hatching, screentone shading, realistic architectural perspective, antique Japanese interior, luminous shoji-window lighting, romantic shojo facial features mixed with horror manga darkness. Keep all text in vertical Japanese manga caption boxes, crisp and legible, with no speech balloons. Constraints: Use exactly 4 panels and exactly 6 text boxes total. Do not add extra characters, extra panels, modern objects, color, sound effects, or watermarks.

Transform this image [Upload Your Image] into a 64K DSLR shot resolution of A creative urban street portrait featuring a young man (same face as the uploaded image) sitting on a curb in a quiet city street, paired with a {argument name="style" default="stylized cartoon illustration"} of himself placed beside him. The real-life subject (same face as the uploaded image) is seated casually on a concrete sidewalk edge, resting his head on one hand in a bored, contemplative pose. He looks slightly upward with a neutral, introspective expression, conveying a sense of quiet thoughtfulness and urban solitude. He wears a {argument name="clothing" default="muted maroon cable-knit sweater"} with a relaxed fit, dark loose-fitting denim jeans with subtle stitched details near the cuffs, and worn brown leather boots. Accessories include layered silver chain dog tag with the name "{argument name="text" default="[Add Text]"}" written on it, rings on his fingers, black-framed glasses, and a light-coloured baseball cap with bold embroidered lettering on the front. His hair is short, dark hair, a well-groomed beard texture and character. Next to him on the sidewalk sits a small {argument name="companion" default="cartoon version"} (same face as the uploaded image) of himself, matching his outfit and pose exactly—same sweater, jeans, boots, cap, glasses, and facial expression. The illustrated character has clean outlines, warm muted colors, and a soft shaded, hand-drawn animation style, creating a playful contrast between realism and illustration. The cartoon figure is proportionally smaller, emphasizing a companion-like or inner-self concept. The background features textured cobblestone pavement, a curb edge, and a building facade with muted yellow walls and green-tiled brickwork near a window. The color palette is earthy and subdued, with natural daylight providing soft, even lighting and realistic shadows. The overall composition blends street photography, fashion portraiture, and mixed-media digital art, evoking themes of identity, self-reflection, and creativity. Style & Mood Keywords: Urban street photography, casual fashion portrait, mixed media art, real and cartoon character duo, introspective mood, muted earthy tones, soft natural light, realistic textures, creative storytelling, high-resolution detail, cinematic realism. Octane render & Unreal Engine