Image-to-Image Restoration of Vintage Photos
A detailed JSON-formatted prompt for the Gemini app's Nano Banana Pro, designed for image-to-image restoration of damaged vintage photographs, focusing on preserving identity and expression while achieving a…

- Prompt included
- Model and format preset
- Free template

The prompt behind this image
Copy the exact prompt behind Image-to-Image Restoration of Vintage Photos, or open it in the studio and make it yours.
- Task
- image-to-image restoration
- Input Image
- output_from_prompt_1
- Reference Image › Type
- face_reference
- Reference Image › Description
- The same face as the damaged photo, keeping the identity unchanged as-is
- Restoration Settings › Restore Face
- true
- Restoration Settings › Restore Skin Texture
- true
- Restoration Settings › Remove Damage
- true
- Restoration Settings › Preserve Expression
- true
- Restoration Settings › Preserve Pose
- true
- Image Settings › Aspect Ratio
- 2:3 portrait
- Image Settings › Resolution
- 8K ultra-HD
- Image Settings › Color Mode
- natural warm color
- Image Settings › Clarity
- high-end modern portrait
- Subject › Pose
- No change from the damaged photo
- Subject › Expression
- The same calm smile
- Subject › Details
- Natural skin texture, realistic eyes, clean hair detail
- Lighting › Type
- soft cinematic studio lighting
- Lighting › Quality
- even and appealing, a modern look
- Background › Style
- A clean, neutral studio background
- Background › Look
- Soft bokeh, no texture damage
- Quality Targets
- no scratches, no creases, no stains, no blurring, no signs of aging/deterioration
- Realism
- A photorealistic modern portrait photo
- Negative Prompt
- changing the face, changing the identity, excessive smoothing, plastic-like skin, an artistic illustration look
Raw JSON
{
"task": "image-to-image restoration",
"input_image": "output_from_prompt_1",
"reference_image": {
"type": "face_reference",
"description": "The same face as the damaged photo, keeping the identity unchanged as-is"
},
"restoration_settings": {
"restore_face": true,
"restore_skin_texture": true,
"remove_damage": true,
"preserve_expression": true,
"preserve_pose": true
},
"image_settings": {
"aspect_ratio": "2:3 portrait",
"resolution": "8K ultra-HD",
"color_mode": "natural warm color",
"clarity": "high-end modern portrait"
},
"subject": {
"pose": "No change from the damaged photo",
"expression": "The same calm smile",
"details": "Natural skin texture, realistic eyes, clean hair detail"
},
"lighting": {
"type": "soft cinematic studio lighting",
"quality": "even and appealing, a modern look"
},
"background": {
"style": "A clean, neutral studio background",
"look": "Soft bokeh, no texture damage"
},
"quality_targets": [
"no scratches",
"no creases",
"no stains",
"no blurring",
"no signs of aging/deterioration"
],
"realism": "A photorealistic modern portrait photo",
"negative_prompt": [
"changing the face",
"changing the identity",
"excessive smoothing",
"plastic-like skin",
"an artistic illustration look"
]
}How to use Image-to-Image Restoration of Vintage Photos in 3 steps
- 1
Open
The prompt is already in the box next to the preview, with the template's model and format selected.
- 2…a portrait of your subject in softgolden-hour light, 85mm…
Personalize
Edit the subject and creative details for your destination.
- 3
GenerateGenerate
Press Generate — your image is created in PicX Studio, so you can compare it with the original and refine one variable at a time.
More like this
Templates with the same subject, style or setup. Open one to see its prompt.
Make it yours
Image-to-Image Restoration of Vintage Photos works as-is. These are the safest ways to change it without losing the look.
Swap the setting
Keep the rest of the prompt and replace the location block. The current setting is “true”.
Swap the light
One lighting change is enough to compare takes. Current light: “soft cinematic studio lighting”.
Change one field
Copy a single group, rewrite it, and generate again so you can see what that field is doing.
| If the result… | Why | Fix |
|---|---|---|
| The result drifts from the preview | Too many fields changed at once. | Reuse Image-to-Image Restoration of Vintage Photos as-is first, then change one group — subject, light, or setting. |
| Nano Banana 2 ignores part of the JSON | Some models flatten or skip nested keys. | Keep the same values and paste them as a short paragraph instead. |
More Nano Banana 2 templates
Other prompts made for Nano Banana 2.
More image prompts
Fresh picks from across the catalog.
Browse template categories
Image-to-Image Restoration of Vintage Photos questions
How do I use the Image-to-Image Restoration of Vintage Photos template?
Add your photo if the template asks for one, edit the prompt if you like — the template's model and format are already selected — and press Generate. Your version opens in PicX Studio.
Can I customize the Image-to-Image Restoration of Vintage Photos template?
Yes. You can change the subject, setting, lighting, mood, format, and available model controls before creating your image.
Is the Image-to-Image Restoration of Vintage Photos template free?
Opening and editing this template is free. Generating uses credits, shown on the Generate button when you are signed in.
Which AI model should I use?
Start with Nano Banana 2, which is the model associated with this template. You can compare another compatible model for a different finish.
Looking for something else? Browse all templates.

Your photo, any look
Thousands of ready templates for every top image and video model, from one credit balance.































