Prompt d'image Branding et UI Ajoutez votre photo
Photo en icône 3D arrondie sur affiche
Une affiche où votre photo reste entière en haut, avec en dessous une icône carrée arrondie pseudo-3D, vue de face, tirée du sujet et de l'ambiance de la photo. L'icône se lit d'un coup d'œil et garde son idée même réduite à la taille d'une application.
- Modèle
- GPT Image 2
- Type
- 3 images
- Format d'image
- 2:3
- Partagé
- 28 août 2026
- J'aime sur X
- 17
Prompt d'image
4,829 caractèresTurn each photo I upload into a separate high-end design poster, no multi-image collage; each photo is output individually. Overall a 3:4 vertical composition, with the top and bottom zones strictly 1:1, each occupying 50% of the frame. The top half keeps the original photo, preserving the subject's identity, structure, pose, real texture, natural light and shadow, and original color atmosphere, with only a slight high-end color grade, giving it the quality of an art magazine, independent publication, and exhibition image. To fit the aspect ratio, the environmental background may be naturally extended, but the subject must not be stretched, distorted, or altered. The bottom half extracts the most recognizable **subject, contour, structure, pose, and narrative relationship** from the photo: first understand the emotion, action, relationship, or visual metaphor truly worth preserving, then reorganize them into a **pseudo-3D rounded-corner icon visual with an original concept**. Don't simply shrink the photo into an icon; instead rebuild the original image's "soul" with the fewest but most essential elements, so the subject is recognized at first glance, the relationship read at the second, and the original scene's emotion felt at the third. The icon's outer contour must always remain a **standard 1:1 square rounded rectangle**, with equal sides, uniform rounding, frontal view, and stable proportions, like a complete, standardized premium App Icon. Elongating, flattening, tilting, perspective distortion, trapezoiding, or letting the subject alter the icon's outer contour is forbidden. The rounded square is only a fixed visual stage; all creativity, space, and narrative happen inside this boundary. The icon's interior uses a **frontal pseudo-3D / 2.5D spatial language**. The subject is the absolute visual core, surrounded by only a few minimal environmental relationships that reinforce the narrative; rich dimensionality is formed through real front-back occlusion, volumetric thickness, height layers, soft facets, local relief, and spatial distance. Multiple elements may have different depths but should grow naturally from the same continuous space, not stacked layer by layer like paper cutouts. Avoid isometric dioramas, top-down models, paper stacking, and complex realistic perspective. The composition prioritizes finding the original image's most powerful **core visual relationship**, and the metaphor can be strengthened through proportion exaggeration, local enlargement, front-back interweaving, form echoing, spatial compression, or symbolic transformation, but all changes must serve one clear theme. The image keeps a clear visual center of gravity, a distinct subject contour, and good negative space, so the icon retains strong recognizability and a unique memory point even when shrunk, rather than an ordinary "object-centered" template icon. The color scheme is **entirely derived from the original photo above, with no preset fixed color menu**. First identify the main color that best represents the subject's identity and emotion, the most vital bright color, and the auxiliary color that carries spatial balance, then remix them while keeping the original hue memory: moderately raise brightness, increase purity, remove dirty gray, organize warm-cool relationships, separate primary-secondary layers, and reinforce the most spirited color. The final colors can be more vivid, more translucent, and more full of life than the original photo, but should still feel as if they originate from it, not applied from another color template. Avoid darkness, aged look, average color sampling, focus-less multicolor, fluorescence, and cheap candy feel. The texture uses a **delicate, matte, soft modern pseudo-3D sculptural feel**, between high-end editorial illustration, miniature art model, and brand 3D visual. Forms have real volume and soft layers but are not overly smooth, reflective, or CG-like; avoid paper-cutout feel, plastic toy feel, glass jelly feel, metallic highlights, and game UI texture. The bottom half's periphery keeps a large clean white space, showing only this one complete rounded-square icon. If text is needed, use only a very small amount of refined editorial text placed outside the icon, establishing a quiet, restrained relationship with the overall layout, without adding peripheral decoration or visual noise. The overall pursuit is **a fixed rounded-square stage + distillation of the original image's soul + frontal pseudo-3D layering + artistic re-direction of the original image's colors + strong recognizability and original metaphor**. The final result should look like a work that could truly enter a brand visual proposal or icon creative review, not a photo thumbnail, cutout illustration, isometric model, or templated App Icon.
Comment l'utiliser
Importez une seule photo ; chaque image est conçue pour être traitée individuellement plutôt qu'en collage. Ce prompt a été partagé comme prompt GPT Image 2, donc choisissez un sujet bien net avec une pose lisible. Conservez le carré arrondi strict en 1:1 et le point de vue frontal, car ces règles donnent au résultat l'allure d'un élément de marque. Vous pouvez librement changer le sujet principal et les indices de palette de la photo, mais gardez le langage des calques et la texture mate.
Partagé publiquement sur X le 28 août 2026. Tout le mérite de ce prompt et de son résultat revient à son créateur. Ce prompt est le vôtre ? Demandez une mise à jour ou un retrait.














