Black and White Photo to Full RGB Color AI Prompts for Image Upscaling by Prompt Craft

Subject : Use reference image as subject from uploaded reference image only, maintaining the exact same person.
Input Condition & Colorization Goal
The uploaded photo is a graduation portrait captured entirely in black and white or grayscale, with no color information present in skin tones, clothing, hair, or background elements. The objective is a full, historically accurate RGB colorization that transforms the monochrome image into a rich, believable color photograph while respecting the era, setting, and likely material tones present in the original scene.
Technical Colorization Process
Analyze grayscale luminance values across skin, hair, fabric, and background to assign historically and physically plausible color values, ensuring skin tones look natural and consistent with the subject's ethnicity and lighting conditions. Assign realistic color to clothing fabrics, accessories, and background elements based on contextual cues such as fabric texture, shadow depth, and period-appropriate material choices. Preserve original grain, contrast, and tonal range from the black-and-white source so the colorization looks integrated into the photograph rather than an artificial digital overlay.
Output Quality & Resolution
Deliver a high-resolution, fully colorized image with smooth, natural color transitions, accurate skin tone rendering, and balanced saturation that avoids oversaturated or cartoonish results. Shadows and highlights should retain their original grayscale depth while now carrying believable hue and chroma, producing a photograph that looks as though it was originally captured in color.
Special Instructions
Preserve the subject's identity with absolute fidelity to the reference image, maintaining 100% facial and anatomical likeness. Keep all facial features, expressions, age, ethnicity, skin tone, hair, and unique details such as moles, scars, freckles, and natural asymmetry exactly unchanged. Retain the original body proportions, posture, and biometric characteristics without unwanted modifications. Do not replace the face with a generic AI appearance or alter the person's true identity in any way. Do not alter facial structure, expression, pose, or composition while adding color; only chroma and hue information should be introduced to the existing grayscale luminance data. The final colorized image should feel authentic and historically respectful, suitable for family archives, museum displays, or digital heritage projects, converting the old image to new image quality without losing its original character.