Adjust photo colours with AI
Develop a photograph in darktable and keep a full-size image alongside the settings used to make it.
Adjust photo colours with AI when a decent photograph needs a change you can put into words. In a catalogue shot, for instance, the glaze on a green bowl may look right while the pale tabletop has turned blue. With photo-xmp-editing you can ask for that cast to be corrected and keep the bowl as a point of comparison.
The adjustments go into a darktable sidecar. You can revisit the edit with the original photograph, or hand over the finished image to someone who only needs it for a listing.
What it does
- Compare a product shot with a reference you choose.
- Lift a subject out of shadow while keeping the light on the background believable.
- Check textured materials at full size before accepting sharper detail.
- Compare recurring colours across a set photographed under different lights.
- Deliver an editable version alongside a full-resolution sRGB export.
Install and use
- Install the skill, Python 3.10 or later, the maker’s photo-xmp tool and darktable. The setup guide also covers utilities for metadata and previews.
- Choose an original photograph. Identify one colour relationship to preserve and one change to make.
- Review the result at ordinary viewing size, then inspect the details that matter before requesting the export and sidecar.
FAQ
Source
Made by f1shh. Original skill: photo-xmp-editing, published on GitHub under the MIT licence.
View on GitHub