Skip to main content

Old photo repair and colorization / ID photo outfit change

Use reference-image editing to remove scratches, colorize, and change outfits while keeping the person's appearance unchanged.

Disclaimer: The content shown on this page is all generated by AI models and is for reference only; it does not constitute any guarantee of results. Actual output and Credits consumption are subject to "Usage Details" in the console.
Upload an old photo or ID photo, and use wan2.7 to repair, colorize, change the background and outfit, and resize — both daily life and job hunting need these. This playbook walks through it with a single old photo: pass the original as a reference image to wan2.7-image, and in one instruction remove scratches and colorize naturally, keeping the person's facial features and expression completely unchanged. Outfit and background changes for ID photos work the same way — one casual photo can be turned into a standard ID photo. Call chain: wan2.7-image reference-image editing (upload the original photo as a reference image).

Actual output

Same input photo, only one repair instruction executed, before/after comparison of repair and colorization:

Before repair (old photo)

Old photo before repair

After repair and colorization

After repair and colorization

Steps: reference-image editing for repair and colorization

Pass the original photo as a reference image to wan2.7-image and send the repair instruction: 
(Attach the old photo as a reference image)
Repair this old photo: remove scratches, folds, and noise, improve clarity,
colorize naturally according to authentic historical colors, keep the person's facial features and expression completely unchanged, do not over-beautify
Outfit and background changes for ID photos work the same way — use a casual photo as a reference image: 
(Attach a casual photo as a reference image)
Process the person into a standard ID photo: formal white shirt with dark suit, solid blue background,
front-facing, uncapped composition, even lighting, keep the person's appearance exactly the same, output in portrait proportions

Notes

  • Consumption: This playbook involves image editing (wan2.7-image with a reference image), repair and colorization 1 time.
  • Boundary: The reference image input is the main cost driver; the higher the original resolution, the more expensive it is.
Get Started
Billing
Token Plan
Model Playground
Clients and Developer Tools
Model Inference
Fine-tuning
Evaluation
Model Compression
Statistics and Monitoring
Model Data
Security and Compliance
Best Practices
Support
Changelog