Photo to Pixel Art Conversion
Prompt by metadigital
5 hari yang lalu

Photo to Pixel Art Conversion

logo Chat GPT
Chat GPT
GPT-5
Transform the uploaded photo into a pixel art character portrait rendered in authentic SNES-era style (Super Nintendo Entertainment System, circa 1990–1997). [STYLE TRANSFORMATION] Convert the subject fully into pixel art using a strict SNES color palette: deep blacks, warm mid-tone shadows, bright saturated highlights — no more than 32 colors per scene, dithering allowed for gradients. Pixel grid size: approximately 16×16 to 48×48 per character sprite tile, scaled up with no anti-aliasing. Hard pixel edges only, no blurring or smoothing. [FACE ANCHOR] Preserve the facial identity of the person in the uploaded photo. Reconstruct facial features — face shape, skin tone, eye shape, hair color — into pixel art form while maintaining recognizable likeness. Expression: free and matching the original photo's mood. Do not alter, replace, or composite a generic face. [ANATOMY LOCK — metadigifun standard] Body proportions: 1:7 head-to-body ratio, natural shoulder width, arms reaching mid-thigh. Each hand: exactly 5 fingers, natural knuckles, 2-joint thumb. No mitten hands, no floating limbs, no merged fingers. Fix only hands or body if anatomy is off — do not alter the face. [BODY & POSE] Recreate the subject's outfit, pose, and expression in pixel art form. Clothing colors must map to the nearest SNES palette equivalent — reduce to flat blocks of color with 2–3 shading steps per garment. Pose matches the uploaded reference. [RENDER RULES] - Hard pixel grid, no anti-aliasing, no sub-pixel rendering - Visible pixel blocks at 100% zoom - SNES color depth: 15-bit BGR, max 256 colors globally / 32 per layer - Background: clean flat color or simple SNES-style gradient background (2–4 solid rows) - No text, no watermark, no HUD, no UI elements, no game title anywhere in the image [FRAMING & RATIO] Output aspect ratio: 4:5 (portrait). Subject centered, full body or 3/4 bust depending on reference. Pixelated border optional but subtle. Generate the image with no text, no watermark, no logo of any kind — three times: no text, no watermark, no logo.

Serupa tapi tak sama