Old Photo Restore — ChatGPT Prompt

Restore and finish this photo for the use case: ChatGPT Prompt for Old Photo Restore.

Context handling:
- First use relevant information already available from the conversation, attachments, or connected sources.
- Do not ask me to repeat details you already have.
- If a required detail is missing, ask one short follow-up.
- State important assumptions. Never invent facts, metrics, employers, customers, logos, credentials, or medical diagnoses. Use [METRIC NEEDED] or [needs proof] when evidence is missing. Educational framing only — not professional, legal, or medical advice.

FACE LOCK (non-negotiable when a face is present):
- Preserve exact facial structure, face shape, eyes, brows, nose, lips, jawline, skin tone, age cues, hairline, and hairstyle.
- Do not beautify, slim, reshape, age, de-age, symmetrize, or replace the face.
- Keep natural pores, fine lines, hair strands, and realistic skin texture. Avoid waxy/plastic skin.

Job: Restore a damaged or faded photograph. Repair defects only. Keep identity.

Defaults:
- Damage / scan notes: {{notes}}
- Color / finish: {{finish}}
- Crop / framing: {{crop}}

Rules:
- Repair scratches, dust, tears, and fade without inventing a new person.
- If colorizing, use historically plausible tones; do not invent uniforms, medals, or jewelry that are not in the source.
- Output a single restored still. Produce the final image only.

Published example

Sample image output

Restored still, same people, fewer defects.

Example context

Example input