chahan

so not real
weekly playlist curator

[chahan @chahandestroiya]

i made a wardrobe website you can use as reference image
all created with nano banana 2.1 so nothing is copyrighted.

⬇️link⬇️

#ai #aigenerated

[chahan @chahandestroiya]

concept:
- "a photorealistic yet dreamlike editorial photo of the person from the reference image standing at the center of a maximalist wall-to-wall installation of art and toys"
- "only the person's face and body shape are carried over from the reference"
- "the dress, hair, accessories, and props are redesigned from the reference image's colors, motifs, patterns, objects, textures, and mood"
- "the entire surrounding world (walls, shelves, furniture, floor, toys, art) is also derived from the reference image"
- "the dress, hair, and surrounding world share the same motifs and the same color family, but the person is clearly separated from the surroundings by brightness and saturation"
- "the walls and the floor are both covered completely, packed to the brim with objects, with no empty wall space and no bare floor visible"
- "the person's silhouette must read instantly against the dense background"
- "must look like a real photograph of a physical set, with a dreamlike atmosphere"

reference_direction:
identity:
- "faithfully preserve the reference person's recognizable facial identity"
- "preserve face shape, jawline, cheek volume, facial proportions, eye shape and spacing, nose, lips, skin tone, distinctive asymmetry, and age"
- "preserve the reference person's body shape, height proportions, shoulder width, and build"
- "no generic beautification, no smoothing or slimming of the face or body"
traits_to_keep:
- "face"
- "body shape"
traits_to_redesign:
- "dress: an elaborate, sculptural dress designed from the analysis of the reference"
- "hair: a striking hairstyle with hair accessories, designed from the analysis of the reference, in colors and shapes drawn from the reference palette"
- "accessories, shoes, and props: designed from the reference's objects and motifs"
- "expression and pose: decided freely to suit the scene, as a striking fashion-editorial presence"
visual_analysis:
- "analyze the reference image's dominant colors, accent colors, fashion, accessories, motifs, patterns, textures, objects, setting, and mood"
- "derive the dress design, hair design, wall decoration, floor decoration, toys, art, furniture, and palette from these elements"
- "small motifs may be enlarged, repeated, abstracted, or transformed into major elements, such as a pattern becoming a dress print, an accessory becoming a wall sculpture, a prop becoming a giant toy"
- "reinterpret meaningful details creatively rather than copying literally"
- "do not introduce unrelated generic motifs"

subject:
type:
- "real human person, photorealistic face and skin with natural texture, visible pores and fine facial hair"
- "realistic human anatomy with the reference's body shape"
- "full body visible, standing at the center of the frame"
fashion:
- "a sculptural, voluminous, highly detailed dress as the centerpiece of the image"
- "print, patchwork, embroidery, trim, and layered ruffles built from the reference's motifs and palette"
- "the dress and hair use the brightest and most saturated versions of the reference palette, so they stand out from the surroundings"
- "the dress has a bold, clean outer silhouette with a clearly readable edge"
- "real wearable fabric with visible weave, natural folds, and stitching, never paper or plastic clothing"
- "matching shoes and accessories in the same visual language"
hair:
- "a bold, designed hairstyle with large hair accessories derived from the reference"
- "real hair strands with natural shine and flyaways"
- "hair color contrasts clearly with the wall behind it"

world:
wall_and_set:
- "walls completely covered, floor to ceiling and corner to corner, with a dense collage of framed art, shelves, ornate cabinets, clocks, mirrors, vases, and objects, with no wall surface visible between items"
- "the art, objects, and ornaments are reinterpretations of the reference's motifs, colors, and objects"
- "toys and collectibles of many sizes fill every shelf and ledge: dolls, plush, figurines, tea sets, and giant props, all drawn from the reference's motifs"
- "an oversized piece of furniture behind or near the person, in the reference color family"
floor:
- "the floor is completely covered edge to edge with objects, leaving only the small spot where the person stands"
- "layered piles and arrangements of toys, plush, tea sets, books, cards, flowers, giant props, rugs, and scattered collectibles, all drawn from the reference's motifs"
- "objects of many sizes overlap each other at different heights, from small items to large standing props"
- "no bare floor, no plain carpet or flat color area visible"
density:
- "hundreds of layered objects on the walls and hundreds more on the floor, with tiny hidden details that reward close looking"
- "every area of the wall and floor is filled, with objects touching and overlapping"
- "clear foreground, middle ground, and background layers"
construction:
- "physically built set with real objects, real depth, and distinct cast shadows between items"
- "some foreground elements partially overlap the person's lower dress hem, but never cover the dress silhouette or the face"
- "a soft contact shadow and a thin shadowed gap around the person's outline, so the dress edge and the objects behind it never merge"
- "objects directly behind the person are slightly darker and less saturated, so the person's outline reads clearly"
- "avoid flat collage, printed backdrop, or wallpaper appearance"
palette:
- "the same color family as the reference, split into two levels"
- "surroundings: one step darker, more muted, and less saturated, with metallic or accent details"
- "person (dress and hair): brighter, cleaner, and more saturated, with a distinct value contrast against the wall behind"
- "do not add new colors outside the reference palette"

atmosphere:
- "dreamlike, surreal, and slightly otherworldly, like a lucid dream built as a real set"
- "soft glow, gentle haze, and saturated yet harmonious colors"
- "the dreamlike quality comes from the set design and color, while the person and materials remain photographic"

camera:
- "front-facing, eye-level, centered, symmetrical composition"
- "full body visible, portrait 4:5 framing"
- "35mm photo look with deep depth of field, the person and the set sharply defined"
- "slight film grain, natural contrast"

lighting:
- "the person is lit by a soft spotlight and a clean rim light that outlines the dress, hair, and shoulders"
- "the surroundings are lit one step dimmer than the person, with distinct shadows between objects"
- "realistic soft light on the skin"
- "dreamy glow kept gentle so it never blurs the person's outline"
- "lighting color adapted to the reference's mood"

critical:
must:
- "reference face and body shape carried over"
- "dress and hair designed from the reference analysis"
- "walls completely filled with art and toys derived from the reference"
- "floor completely filled with objects and toys derived from the reference"
- "extremely dense, layered, physically built set with no empty wall and no bare floor"
- "person's silhouette clearly separated from the background by brightness, saturation, rim light, and contact shadow"
- "same motifs and color family across dress, hair, and world"
- "photorealistic person and materials with a dreamlike atmosphere"

avoid:
- "identity drift"
- "generic beautification"
- "changed face shape or body shape compared to the reference"
- "copying the original outfit or hair literally"
- "dress and hair blending into the background, merging with wall objects, or a muddy silhouette"
- "dress with the same value and saturation as the wall behind it"
- "paper, plastic, or toy-like clothing"
- "airbrushed, porcelain, or overly smooth skin"
- "chibi, doll-like, or cartoon person"
- "empty or sparse wall areas"
- "bare floor, plain carpet, empty floor space, or a flat color floor"
- "flat collage, printed backdrop, or wallpaper appearance"
- "figure that looks pasted in or lit differently from the scene"
- "generic decorations unrelated to the reference"
- "garbled text or extra text on the image"
- "watermarks or brand logos"

[chahan @chahandestroiya]

Drop in one reference photo → a dreamlike maximalist set built from it.
Only your face and body shape carry over. The dress, hair, wall art, toys, and floor are all redesigned from the photo’s colors and motifs, packed wall to floor with no empty space. 🧸🖼️

⬇️prompt⬇️

#AIart #ai #aiprompt #promptshare

[chahan @chahandestroiya]

@EEJ_OOXO 액션이 멋져서 몇번이고 돌려봤어요!

[chahan @chahandestroiya]

the figure's style changes depend on the mood and tone of your model!

[chahan @chahandestroiya]

Drop in one reference photo + a name →
a limited edition figure box set.

The prompt reads the photo and decides the box color, title, Japanese copy, and printed illustrations on its own. Only the hair, outfit, accessories, and expression carry over.

⬇️prompt⬇️

#AIart #promptshare

[chura🪽 @churababy]

<= gpt 2.5 sunburst
nanobanana 2.1=>

[chahan @chahandestroiya]

input:
name: "{NAME}"

concept:
- "a photorealistic product photo of a Japanese-style limited edition figure box set, with a small PVC figure visible through the clear window of the box"
- "the figure is an original stylized character designed from the reference image, not a human likeness"
- "the box design, color, text, logos, and printed illustrations are all decided from the reference image"
- "the box, window, figure, and lighting are all real-photo rendered; the printed illustrations on the box are 2D printed artwork"

reference_direction:
traits_to_transfer:
- "hair color, hairstyle, bangs, and hair accessories"
- "outfit colors, outfit silhouette, and accessories"
- "props held or worn"
- "facial expression and mood, translated into the figure's face"
- "overall vibe and fashion identity"
not_transferred:
- "the face is fully stylized in the figure's art style, decided freely"
- "do not replicate the reference person's realistic face, skin, or facial structure"

visual_analysis:
- "analyze the reference image's dominant colors, accent colors, fashion, accessories, motifs, and mood"
- "decide the box color scheme, product concept, title, subtitle, catchphrases, and edition label on your own from this analysis"
- "reinterpret details creatively rather than copying literally"
- "do not introduce unrelated generic motifs"

figure:
style:
- "small collectible PVC figure with a painted finish, in a cute Japanese stylized character art style"
- "large head with a simple stylized face, big eyes, small simplified nose and mouth"
- "slender body with thin, long arms and legs, large boots or shoes"
- "solid flat paint colors with soft plastic highlights"
design:
- "hair color and hairstyle from the reference sculpted as a solid-painted hair shape"
- "outfit rebuilt from the reference's clothing colors and silhouette in simple painted form"
- "accessories and props from the reference as small sculpted items"
- "expression from the reference translated into the stylized face"
pose:
- "standing upright on a small base, facing forward, with a small characteristic gesture matching the reference's mood"
- "visible through the clear window in the center of the box"

illustration_style:
era_and_medium:
- "printed illustration in the style of a late 2000s Japanese rhythm game package"
- "simple, cute, pop character art with a handheld-game-box feel, not a modern anime or key visual style"
linework_and_color:
- "clean uniform-weight outlines with flat cel-style coloring"
- "almost no gradients, no soft airbrush shading, no detailed highlights"
- "bright pastel and primary colors on solid color backgrounds"
character_drawing:
- "large round simple eyes with minimal highlights, no detailed iris, no heavy eyelashes"
- "nose and mouth reduced to tiny dots or short lines"
body_proportions:
- "about 2.5 to 3 heads tall, a stylized chibi-leaning pop character, not realistic or semi-realistic proportions"
- "head is large and round, roughly one third of the total height"
- "neck is short or barely visible"
- "torso is small, short, and slim, with minimal chest or waist definition"
- "arms are thin and long relative to the torso, with simple tapered shapes and small simplified hands (mitten-like or with minimal finger lines)"
- "legs are slim and fairly long relative to the torso, with straight simple lines and little muscle or thigh detail"
- "feet and shoes are oversized and chunky (big boots or sneakers), clearly larger than the hands"
- "overall silhouette is a big head on a slender stick-like body with big shoes"
posing:
- "lively, dynamic idol-style poses: one arm raised or stretched out, legs slightly crossed or in mid-step, body tilted with energy"
- "limbs drawn with smooth, simple, flowing lines"
side_panels:
- "side strip panels are separate framed cells, each with a solid pastel background color (yellow, mint, light green, light blue) and a simple line-art character illustration"
decoration:
- "simple star, heart, and music note icons, rounded handwritten-style logo lettering"

box:
layout:
- "upright Japanese game-package style box with a large clear window in the center showing the figure"
- "top: a fictional title logo and a fictional Japanese subtitle"
- "side or lateral strips: small printed illustration panels of the character, in the style defined in illustration_style"
- "one side: a larger printed illustration of the character in the style defined in illustration_style, with a limited edition label such as a figure-included limited edition badge"
- "bottom: a dark band with a fictional publisher logo and fictional rating and age marks"
- "the character name from the input appears on the box as the character name"
color:
- "box main color, accent colors, and band colors decided from the reference's palette and mood"
text:
- "name: use the input name exactly as written at the top"
- "all other box text is decided from the reference, in natural Japanese box-package style, with some small English text mixed in"
- "all logos are fictional original designs"
- "all text is crisp, correctly spelled, and fully legible"

camera:
- "product photo of the box, slightly angled three-quarter view, centered, box fills the frame"
- "clean white or light neutral background"
- "sharp focus on the box and figure"

lighting:
- "soft studio lighting with realistic reflections on the clear window and glossy printed surfaces"
- "soft shadow beneath the box"

critical:
must:
- "photorealistic product photo of a real box and real PVC figure"
- "figure with large head, slender long limbs, and a simple stylized face"
- "hair color, hairstyle, outfit, accessories, props, and expression carried over from the reference"
- "box illustrations in the late 2000s flat cel-style pop character art and body proportions defined in illustration_style"
- "box color and text decided from the reference"
- "input name rendered exactly and legibly"
- "all logos fictional"

avoid:
- "realistic human face on the figure"
- "figure that looks like a doll, plush, or real person"
- "real brand logos or real console logos"
- "garbled, misspelled, or unreadable text"
- "flat 2D illustration instead of a product photo"
- "generic decorations unrelated to the reference"
- "watermarks"
- "modern anime key visual style, detailed eyes, glossy shading, or gradient coloring on the box illustrations"
- "realistic or semi-realistic proportions on the printed illustrations"
- "long realistic legs, tall slim fashion-illustration proportions, or 5+ heads tall characters on the box illustrations"
- "large detailed hands, realistic feet, or visible muscle and body contour detail"

[chahan @chahandestroiya]

don't forget to put your name at the top of the prompt!

[chahan @chahandestroiya]

Drop in one reference photo and get a live-streaming app screenshot.
The prompt reads the image to decide the mood, colors, stream concept, event title, like-goal tiers, and viewer chat on its own, while keeping the person’s face intact. 📱✨

⬇️prompt⬇️

[chahan @chahandestroiya]

concept:
- "a vertical live-streaming app screen capture featuring a real human streamer, shown mid-broadcast"
- "the streamer remains fully photorealistic and human"
- "the app UI overlay, live concept, title, event, and chat are all derived from the reference image"
- "must look like a genuine screenshot of a mobile live-streaming app, not a poster or illustration"

reference_direction:
identity:
- "faithfully preserve the reference person's recognizable facial identity"
- "preserve face shape, facial proportions, eyes, nose, lips, jawline, skin tone, distinctive asymmetry, age, and core hair silhouette and hair color"
- "no generic beautification or facial reinterpretation"

visual_analysis:
- "analyze the reference image's dominant colors, accent colors, fashion, accessories, motifs, textures, objects, setting, pose, and overall mood"
- "decide the live-stream concept, event theme, and title on your own from this analysis"
- "derive the UI color theme, title wording, event content, and room decoration from the reference's palette, motifs, and mood"
- "reinterpret meaningful details creatively rather than copying literally"
- "do not introduce unrelated generic motifs"

ui_overlay:
style:
- "authentic mobile live-streaming app interface"
- "semi-transparent glassy panels, rounded corners, soft drop shadows"
- "UI accent color taken from the reference's strongest accent color"
elements:
top_left:
- "circular streamer avatar (cropped from the same person) + streamer name + small emoji matching the reference's motif"
- "LIVE badge in the accent color + viewer count pill"
top_right:
- "close (X) and more (...) icons"
- "a rounded like-counter panel with a low current number, plus a tier list of 3-4 goals, each with a short reward label"
left_title:
- "a large stylized event title with thick outline, gradient fill, and glow, in the accent color"
- "a smaller one-line subtitle explaining how likes trigger the event"
- "the title and event concept are decided from the reference's outfit, motif, and mood"
chat_feed:
- "5-7 viewer comments stacked in the lower left, each with a small round avatar"
- "comments fit the concept and the reference's mood: reactions to the look, the title, the event, cheering for likes"
- "every avatar is a simple flat icon inside a small circle, never a photo of a face"
- "avatar types: default gray silhouette placeholder, solid pastel color circle with a single initial letter, or a tiny simple object icon (flower, cat paw, star, food, sky)"
- "avatars are clean vector-style, uniform size, perfectly circular, with no facial features at all"
bottom_bar:
- "comment input field with placeholder text"
- "more (...) button, gift box button, and a large circular heart button in the accent color"
floating_hearts:
- "translucent hearts in 2-3 sizes rising from the bottom right, in accent color shades"
language:
- "all UI text in natural Japanese (change here if desired)"
- "all text must be rendered crisp, correctly spelled, and fully legible with no garbled characters"

world:
setting:
- "a lived-in room or location chosen from the reference's mood and setting"
- "background elements and props reuse the reference's colors and motifs"
- "practical lights creating soft bokeh"
- "shallow depth of field, background gently blurred"

subject:
type:
- "real human person"
- "photorealistic face and skin with natural texture"
- "realistic human anatomy"
fashion:
- "follow the reference's outfit identity and styling"
- "real wearable fabric with visible texture"
- "outfit, room, and UI share one coherent color language"
pose:
- "pose and framing follow the reference image"
- "eyes looking into the camera as if interacting with viewers"
- "natural human proportions"

camera:
- "portrait 3:4 or 9:16 smartphone front-camera framing"
- "subject sharp, background softly blurred"
- "slight phone-camera softness, realistic skin highlights"

lighting:
- "frontal ring-light style key light on the face"
- "ambient glow from background lights"
- "lighting color adapted to the reference's mood"

critical:
must:
- "reference facial identity"
- "real human appearance"
- "complete, believable live-app UI with every element listed above"
- "reference-derived colors, title concept, and chat content"
- "legible, correctly rendered text"
- "chat avatars as clean faceless flat icons"

avoid:
- "illustrated, anime, or cartoon face"
- "identity drift"
- "generic beautification"
- "poster or advertisement look instead of an app screenshot"
- "garbled, misspelled, or unreadable text"
- "generic decorations unrelated to the reference"
- "watermarks or real brand logos"
- "photographic, realistic, or illustrated face avatars in the chat feed"
- "distorted, melted, overlapping, or duplicated faces in small avatars"

[chahan @chahandestroiya]

night stroll is always worth it

[chahan @chahandestroiya]

supports mobile, but mobile version is a bit clanky right now. will update it as soon as possible!

[chahan @chahandestroiya]

i made a wardrobe website you can use as reference image
all created with nano banana 2.1 so nothing is copyrighted.

⬇️link⬇️

#ai #aigenerated

[chahan @chahandestroiya]

https://t.co/7E6KQmti0W

let me know if you want me to add more!

[chahan @chahandestroiya]

trying frutiger aero style :0

[chahan @chahandestroiya]

after workout
chilly night tho

[chahan @chahandestroiya]

new playlists will be uploaded twice every week
<3
this time, music for that moment
when your heart skips a beat!

[chahan @chahandestroiya]

@bingsome 그치만 너무 사랑스러워요

[chahan @chahandestroiya]

@from_rone love your hairstyles so much!!!
and matching nails too!