10 Common Photo Restoration Mistakes (And How to Avoid Them)
Most bad photo restoration results come from a handful of predictable mistakes. Here's what people get wrong and how to fix it.
Why Restorations Go Wrong
Modern AI photo restoration produces excellent results when used correctly. Most bad outcomes come from user mistakes, not tool limitations. This guide covers the 10 most common mistakes and exactly how to avoid them.
1. Applying Colorization Before Cleanup
The mistake: Running Colorize on a scratched or stained photo. The AI colorizes the scratches along with the photo, making the damage more visible.
The fix: Always apply Remove Scratches first (5 coins), then Colorize (4 coins). Clean up physical damage before adding color information.
Why it happens: Colorization is the most exciting feature, so users often apply it first without thinking about sequence.
2. Shooting Captures with Flash
The mistake: Using iPhone flash to photograph an old print, creating harsh reflections and flattened lighting.
The fix: Disable flash. Use indirect daylight only. The photographing old photos guide covers correct technique.
Why it happens: Indoor low-light triggers auto-flash in iPhone Camera app. Always manually disable.
3. Over-Restoring a Photo
The mistake: Running every feature on every photo. A mild-fade photo that just needs Enhance Details gets Remove Scratches + Restore Faces + Colorize + Recreate applied, producing an unnatural oversaturated "plastic" look.
The fix: Match features to actual damage. A photo with no scratches doesn't need Remove Scratches. A color photo doesn't need Colorize. Apply only what the photo actually needs.
Why it happens: More features feels like more effort, which feels like better results. It usually isn't.
4. Using Free Tools for Important Projects
The mistake: Restoring heirloom photos with free tools that add watermarks, low-resolution exports, or aggressive automatic enhancements.
The fix: For any photo that matters (memorial albums, gifts, wall prints), use a paid tool with full resolution and control. Restory's 50-coin pack at EUR 7.99 costs less than a cup of coffee and gives you 4-6 fully restored photos.
Why it happens: Budget constraint. But the cost of a tool isn't meaningful at small scales — the cost of a ruined heirloom photo is.
5. Not Making Backups Before Restoring
The mistake: Overwriting the original scan with the restored version. Later realizing the restoration was wrong and having no path back.
The fix: Keep every version. Save restored photos as new files (Restory does this automatically). Maintain your original scans in a separate folder. Storage is free.
Why it happens: Good intent to "keep things clean." But old photos are irreplaceable.
6. Trusting AI Colorization as Historical Accuracy
The mistake: Presenting an AI-colorized photo as "how it really looked" to family. The AI picks plausible colors, not actual historical colors.
The fix: Be honest about what colorization is — educated guess, not recovered truth. Label AI-colorized versions if you publish them. For family use, the plausibility is usually enough; just don't overclaim accuracy.
Why it happens: Colorized photos look so convincing that people forget they're AI predictions.
7. Scanning at Low Resolution
The mistake: Capturing an old photo at low quality because "it's just for digital use." Later wanting to print it and discovering there's not enough resolution.
The fix: Always capture at maximum resolution. Modern iPhones produce 12-48 MP images regardless of planned use. Storage is cheap; re-scanning is tedious.
Why it happens: Zoom-in captures in the Camera app to "fill the frame" effectively reduce resolution. Always pull back and crop in post-processing instead.
8. Shooting Through Glass in Frames
The mistake: Photographing framed photos through the glass without removing them. Glass creates reflections and reduces sharpness.
The fix: Remove the photo from the frame if possible. For valuable antique frames or photos mounted permanently, capture at a very slight angle to eliminate reflections, knowing the result will be slightly degraded.
Why it happens: Effort and sometimes fear of damaging valuable frames. For important photos, the effort is worth it.
9. Restoring in the Wrong Order
The mistake: Apply features in an order that doesn't match the damage. For example, running Enhance before Restore Faces on a portrait, which enhances blurred faces that could have been reconstructed first.
The fix: Follow the canonical order for typical family photos:
- Remove Scratches (physical damage first)
- Restore Faces (if portraits)
- Enhance Details (general sharpness)
- Colorize (only if B&W)
Why it happens: Nobody explains the order, so users experiment randomly. The ultimate guide to photo restoration covers this sequence.
10. Not Asking Family Before Restoring Heirlooms
The mistake: Aggressively restoring or colorizing family heirloom photos without consulting other family members who may have strong feelings about the originals.
The fix: For any photo that matters to multiple family members, ask before restoring. Some people prefer original B&W over colorized. Some find heavy restoration inappropriate for heirloom photos. A simple "I'm thinking of restoring grandma's wedding photo — any thoughts?" prevents conflict and often produces better decisions.
Why it happens: Restoration feels like a clearly-good improvement. But family archives are shared objects with shared meaning.
Bonus: The Meta-Mistake
The mistake: Expecting perfection. Old photos have real limits — badly damaged areas can't be perfectly recovered, very small source images can't be upscaled infinitely, and faces the AI has never seen can't be reconstructed accurately.
The fix: Set expectations. AI restoration produces excellent results in 85-95% of cases, good results in another 4-10%, and imperfect results in the remaining 1-5%. The bad cases are usually where the original data is fundamentally insufficient. Know when to stop.
Why it happens: AI is impressive, so users expect it to be perfect. It's not magic.
The Core Principle
Good photo restoration is about working with what you have, not forcing what isn't there. Start with clean captures, apply features in order, match effort to the photo's actual damage, back up originals, and accept imperfection where it exists. Follow those five rules and 95% of your restorations will exceed your expectations.
Related Reading
- The ultimate guide to photo restoration
- How to photograph old photos
- How to colorize black and white photos
- Restory vs Remini comparison
Frequently Asked Questions
What's the single most common photo restoration mistake?
Applying Colorize before scratch removal. It colorizes the scratches along with the photo, making damage more obvious rather than hiding it. Always clean up physical damage first, then add color. This single change improves restoration quality for the majority of old photos more than any other adjustment.
Can I undo a bad restoration?
If you still have the original scan, yes — just re-restore from scratch. This is why keeping the original scan is critical. If you only kept the restored version and it's wrong, you'd need to rescan the physical print (if you have it) or accept the imperfect result. Maintain multiple versions of every restoration: original scan, restored, and any variations you tried.
Is it better to over-restore or under-restore a photo?
Under-restore. A photo that still shows minor age characteristics looks natural and historical. An over-restored photo looks artificial and often triggers the uncanny valley response — viewers sense something is off without being able to articulate why. When in doubt, apply one fewer feature. You can always run more restoration later; reversing over-restoration requires starting from the original scan.
Restore your photos with Restory
AI to colorize, repair, and animate old photos. 32 languages, free trial.
Try Restory
