Scanned archives are different from fresh photos. Paper grain, scanner noise, faded ink and uneven light are all part of the file. Good old scanned photo restoration starts before any AI tool, with a clean scan and a consistent set of settings that you can apply across a whole box of prints without making every face look new.
The goal for an archive is consistency. One photo should not look dramatically sharper or warmer than the next, unless its original condition truly differs. A repeatable workflow saves time and keeps a family or organisation archive believable.
Prepare the scan first
Scan at a resolution that preserves detail without creating huge files for no benefit. Clean the scanner glass, align the print straight, and save a master copy in a lossless format. Name files clearly, with dates or people where known, so restored versions can always be traced back to the original scan.
Check a few representative scans before batching anything. A setting that works for a clean studio portrait may be too strong for a grainy snapshot. Test on a light photo, a dark photo and a damaged photo, then choose middle settings that work for all three.
Choose AI settings that respect the original
AI enhance can reduce noise and improve clarity, but scanned photos need restraint. Strong enhancement can turn paper grain into strange patterns and smooth skin into plastic. Start with a low or moderate setting, view faces at normal size, and only increase it if detail is genuinely clearer without changing the person.
Sharpening should follow the same rule. A small amount helps text, glasses and clothing, while too much creates halos around edges. For the underlying tools, see the guide to using AI enhance in Photomator. For larger batches, the guide to batch editing in Photomator helps keep quality steady.
Colour, contrast and batches
Old scans often need gentle contrast rather than strong colour. Restore white and black points first, then correct any yellow or magenta cast in small steps. Avoid a single preset for every scan, because paper ages differently. Instead, use a base preset for exposure and noise, then adjust colour per photo when needed.
When working in batches, process in small groups and review the results before moving on. If a setting starts to drift, stop and reset it. A consistent archive is more valuable than a fast one, and corrections are much harder after files have been exported and shared.
Frequently asked questions
What scan resolution is best for restoration?
For most standard prints, 600 dots per inch gives a good balance of detail and file size. Small prints benefit from higher resolution, while very large prints can use a lower setting without losing useful detail.
Should AI enhance be applied before or after colour correction?
Correct major colour and exposure problems first, so the AI tool works on a balanced image. Apply enhancement lightly, then check colour again, because sharpening and noise reduction can change how colour appears.
How do I avoid faces looking artificial?
Use moderate enhance and sharpening settings, and compare faces with the original scan. If eyes, wrinkles or glasses look altered, reduce the effect. Recognition should always come before apparent sharpness.
Can I batch restore a whole archive?
Yes, but in small, reviewed batches. Use consistent base settings, then make small individual adjustments. Review each batch at normal viewing size before exporting.
What file format should I keep for restored scans?
Keep a lossless master copy of the original scan and export restored versions separately. That way the archive remains safe even if a later edit or export choice does not work out.
Build a steady archive workflow
Old scanned photo restoration works best when it is calm and repeatable. Prepare clean scans, test settings on a few examples, keep AI effects moderate, and review in batches. The result is an archive that looks clearer and more consistent, while still feeling like the photographs it came from.