Bring black-and-white photos back to colour.

How AI colourisation works

The editor uses DDColor, a modern colourisation network (Apache-2.0 licence) that downloads to your browser the first time you use it and then runs on your own device. It looks at the whole photo at once — recognising sky, skin, foliage, water, stone and fabric — and predicts the colour channels for each area. Those colours are then combined with your photo's original lightness at full resolution, so no sharpness or detail is lost.

A Colour intensity slider lets you go from a subtle, hand-tinted look to rich, vivid colour.

What to expect — honestly

• Natural scenes look most convincing: skies, sunsets, water, grass, trees and skin tones.
• Clothes, cars and painted objects are a guess. A black-and-white photo doesn't record what colour a dress or a door was, so the AI chooses a plausible colour. If you know the real colours, treat the result as a starting point.
• Very damaged prints colourise better after restoration — use Restore Old Photos first, which can colourise as its final step.

Tips for the best result

• Scan prints at 600 DPI or higher — see our guide to scanning and upscaling old photos.
• Crop away scanner borders and white frames first so they don't pick up colour.
• Lower the intensity slightly for portraits; skin looks more natural at 70–90%.
• Finish with Auto Enhance or a gentle filter in the same editor.