Photogpt
Restore the uploaded image into a clean, fully reconstructed, high-quality photograph while preserving the original people, pose, composition, and scene. ⸻ Goal Transform the damaged, degraded, or old photo into a modern, clear, natural-looking image while keeping the identity and scene unchanged. The output should look like the same photo taken with a modern camera, not an altered or reinterpreted scene. ⸻ Identity Preservation Strictly preserve: • facial structure and proportions • eye shape and spacing • nose structure • lip shape • head shape • hairstyle and hairline • age characteristics • clothing style and form • body pose and gesture • relative position of people in the frame Do not redesign the people. ⸻ Restoration Tasks Perform advanced restoration: • remove scratches • remove stains • remove cracks and dust • reconstruct missing areas • repair torn regions • repair faded textures • remove film damage and compression artifacts Rebuild lost facial details naturally while maintaining identity. ⸻ Image Enhancement Improve overall image quality: • natural skin texture • realistic facial details • balanced exposure • corrected contrast • natural colors (if colorized) • improved sharpness without artificial smoothing The image should appear clean and naturally photographed. ⸻ Colorization If the image is black and white: • generate natural realistic color tones • natural skin color • realistic hair color • believable clothing colors • subtle film-style color palette Color must appear authentic and historically plausible, not oversaturated. ⸻ Lighting Reconstruct natural lighting consistent with the original scene. Avoid artificial studio lighting. Maintain the direction and softness of the original light. ⸻ Texture Reconstruction Restore realistic textures for: • skin • hair • fabric • background materials • furniture and environment Avoid plastic skin or over-smoothed surfaces. ⸻ Composition Preserve exactly: • camera angle • subject placement • framing • background layout Do not crop or reposition elements. ⸻ Output Quality • ultra clean restored photograph • realistic human skin • sharp facial details • natural photographic look • modern high-resolution clarity ⸻ Constraints No AI beauty filters No face redesign No cartoon style No painterly effects No artificial sharpening halos No unrealistic colors No identity change The output must look like a perfectly restored version of the original photograph.

PhotoGPTのAI画像

Gemini Nano Banana 2.1
プロンプト

Restore the uploaded image into a clean, fully reconstructed, high-quality photograph while preserving the original people, pose, composition, and scene. ⸻ Goal Transform the damaged, degraded, or old photo into a modern, clear, natural-looking image while keeping the identity and scene unchanged. The output should look like the same photo taken with a modern camera, not an altered or reinterpreted…

サイズ1792px x 2240px
参照画像
元画像 1

創作を始める

数千人のアクティブユーザーが集まるDiscordコミュニティに参加して、最新のプロンプトアイデアやコミュニティ作品をチェックしましょう

画像を生成Discordに参加
Google Veo 3.1Seedance 2

関連作品

コミュニティの似たAI生成作品を探索できます

このスタイルを試す