imageFree

Old Photo Restoration

Restores a damaged vintage child portrait into a clean, realistic color snapshot while preserving identity, pose, clothing, and background. (free to use) — Free AI image template on PicX Studio.

gpt-image-2

Sample Prompt

Using the provided reference image, refine and restore the old damaged photo into a clean, realistic original snapshot. Remove all cracks, creases, scratches, faded discoloration, stains, and paper damage, and reconstruct any missing or obscured areas naturally. Preserve the same child, facial identity, joyful laughing expression, hairstyle with pigtails, yellow flowers beside both ears, checkered floral dress with white collar, pose, close portrait framing, and outdoor background. Convert the aged pinkish tint into natural color, sharpen details, improve lighting and contrast, and make the result look like an authentic modern film photograph rather than an illustration. Do not change the subject’s age, clothing design, composition, or expression; do not add extra people, text, borders, or watermark.

Reference assets included

Old Photo Restoration needs input media to work. Open it in the studio and these load alongside the prompt — pick whichever ones your chosen model can take.

1 image

1 reference image

A practical guide to Old Photo Restoration

Use Old Photo Restoration to plan an AI image around gpt-image-2 generation, themes such as en and gpt-image-2. Match the preview's composition first, then personalize the subject, setting, and mood for your own project.

Model direction: start with gpt-image-2, the model this template was tuned for, before comparing another model.
Composition: choose an aspect ratio for the final placement before you generate the image.
Creative direction: use en, gpt-image-2, twitter as the main visual anchors and change one variable at a time.