How to Digitize Old Photos at Home: Flatbed Scanners, Phone Apps, and Getting the Best Scan Quality

Digitizing old family photographs at home is a straightforward project with the right approach, but the quality of the scan directly determines the quality of any AI restoration you apply afterward. This guide covers the practical choices: what scanning resolution to use, whether a flatbed scanner or phone app is sufficient, how to prepare photos before scanning, and how to save files so they are ready for restoration.

What Equipment Do You Actually Need to Digitize Photos at Home?

The minimum equipment for digitizing old photos at home is a smartphone with a good camera. The recommended equipment is a flatbed scanner. The gap between these two options matters, but it is smaller than many guides suggest, and it shrinks further when AI restoration is the next step.

A flatbed scanner in the $80-$150 range — the Epson Perfection V39 or Canon CanoScan LIDE series — produces scans that are significantly better than phone photos for fine detail and lighting consistency. Both scanners produce 1200 DPI optical resolution, which is more than sufficient for standard prints. The software bundled with these scanners is adequate for basic scanning tasks.

If you do not have a scanner and prefer not to buy one, phone-based scanning apps are a practical alternative. Microsoft Lens (free, iOS and Android) applies perspective correction and captures at your phone camera's full resolution. Google PhotoScan (free) uses a multi-frame capture method designed specifically for reflective prints. Both produce workable input for AI restoration — not as clean as flatbed scans, but sufficient to achieve meaningful improvement.

What Resolution Settings Should You Use?

The resolution setting directly determines how much detail is captured from the original print and how much the AI restoration has to work with.

For standard 4x6 prints: scan at 600 DPI. This produces a 2400x3600 pixel file — enough detail for excellent screen display and printing at up to 8x12 inches. For wallet-size photos (approximately 2x3 inches): scan at 1200 DPI to capture equivalent relative detail. For large prints (8x10 or larger): 300-400 DPI is sufficient because the physical print already has more area.

Scanning at higher resolution than these recommendations produces diminishing returns — most flatbed scanners have optical limits around 1200-2400 DPI, and scanning beyond their optical maximum interpolates resolution without adding real detail. The numbers in scanner software that go up to 9600 DPI are marketing figures for interpolated resolution, not optical quality.

How Should You Prepare Old Photos Before Scanning?

Before placing a photo on the scanner, clean both the photo surface and the scanner glass. Use compressed air to blow loose dust off the photo, then wipe gently with a lint-free lens cloth. Never use moisture, alcohol, or cleaning agents directly on a photograph's surface.

Place the photo face-down on the scanner glass with the image aligned to the edge guides. Ensure the scanner lid closes fully over the photo — if a photo is warped or bowed, apply gentle pressure through a soft cloth weight to flatten it against the glass during the scan, which significantly reduces blur from surface-to-glass distance variation.

Scan the full glass area and crop precisely in software afterward rather than trying to position the photo exactly during scanning. This ensures you capture the full image area without accidentally clipping edges.

What File Format and Color Mode Should You Choose?

Color mode: scan color photographs in RGB color even if you plan to restore black-and-white photos — the color channel information helps AI models better characterize the damage type (yellowing versus neutral fading). For true black-and-white photographs (silver gelatin prints), grayscale mode is fine, but RGB is also acceptable.

File format: TIFF for archival storage, JPEG at 90% quality or higher for anything you plan to upload to an online restoration tool. The file size difference is significant — a 600 DPI TIFF runs 40-50 MB versus 3-5 MB for a comparable JPEG — but the quality difference is minimal for AI restoration input. Tools like ArtImageHub, which apply Real-ESRGAN and NAFNet for restoration, work effectively with high-quality JPEG input.

How Does Scan Quality Affect AI Restoration Results?

AI restoration models like Real-ESRGAN and GFPGAN analyze the existing pixel data in your scan and reconstruct what the original image should look like. A clean, high-resolution scan gives these models more accurate and detailed input, which produces better output.

Specific scan quality factors that affect restoration results: lighting consistency (even lighting across the photo surface gives the AI accurate tonal information), resolution (more pixels mean finer detail reconstruction), and alignment (a straight, flat scan without perspective distortion means the AI addresses only the original damage, not artifacts introduced by the scanning process).

After scanning, upload to ArtImageHub for AI restoration. The $4.99 one-time fee covers full-resolution processing through the complete restoration pipeline. The cleaner your input scan, the cleaner the restored output.

Frequently Asked Questions

What resolution should I scan old photos at home?

The correct scanning resolution depends on the size of the original print and what you plan to do with the digital file. For standard 4x6 prints that you want to display on screen or print at the same size, 600 DPI produces a 2400x3600 pixel image — plenty of detail for most uses. For smaller prints (wallet-size, 2x3 photo booth strips), scan at 1200 DPI to capture proportionally equivalent detail. For large prints (8x10 or larger), 300 DPI is usually sufficient because the original has more physical area. For any photograph you plan to restore using AI tools — which benefit from having more pixel data to analyze — scan at 600 DPI minimum. Real-ESRGAN and GFPGAN can extract and reconstruct more detail from a 600 DPI scan than a 300 DPI scan of the same print because there is literally more information in the input file. If you are using a phone scanning app like Microsoft Lens or the iPhone document scanner, capture in the highest resolution the app supports. Phone apps cannot match flatbed scanner quality for fine detail, but higher capture resolution reduces the quality gap.

Are phone scanning apps good enough for AI photo restoration?

Phone scanning apps — including Microsoft Lens, Google PhotoScan, and the built-in document scanners on iOS and Android — are genuinely useful for digitizing photos when you do not have access to a flatbed scanner, but they have specific limitations that affect restoration quality. The primary problems are lighting consistency and geometric accuracy. A flatbed scanner uses uniform, calibrated cold lighting that eliminates shadows and reflections entirely. A phone camera depends on whatever ambient light is available, which typically creates a center-bright gradient and potential reflections on glossy print surfaces. Google PhotoScan was specifically designed for this problem — it captures multiple overlapping frames from different positions and stitches them to eliminate reflections, which works reasonably well for glossy prints. Microsoft Lens applies perspective correction to reduce keystoning. Even with these improvements, phone apps produce scans with more lighting variation and slightly less fine detail than a good flatbed scanner. For AI restoration, the quality difference is real but not disqualifying — AI tools like ArtImageHub can work with phone-captured images and still produce significant improvements, just not quite the same output quality as a clean flatbed scan.

Should I clean old photos before scanning?

Yes, cleaning photos before scanning is worth doing carefully. Dust, fingerprints, and loose debris show up clearly in scans and add extra work for both manual retouching and AI restoration. Use a soft lint-free cloth — the type used for camera lenses or eyeglasses — to gently wipe the scan surface of the photo. Use a can of compressed air to blow dust off the surface before placing the photo on the scanner glass. Never use water, alcohol, or cleaning sprays directly on a photograph, as these can damage the emulsion surface or cause irreversible staining. For photographs with mold or mildew, do not attempt to clean them yourself — the spores can spread and cause further damage, and photos with active mold should be handled by a professional conservator before scanning. The scanner glass itself should also be clean: fingerprints and smudges on the glass appear as artifacts in the scan. Clean the scanner glass with a lint-free cloth before each scanning session. These simple steps reduce the total amount of damage visible in the scan and make AI restoration produce cleaner output, since the AI does not have to account for dust artifacts that were not in the original photograph.

What file format should I save scans in — JPEG or TIFF?

For archival scans of irreplaceable photographs, TIFF is the right format. TIFF stores pixel data without compression, which means every detail captured by the scanner is preserved exactly in the file. JPEG uses lossy compression, which discards some fine detail to reduce file size — and each time you open and resave a JPEG, additional quality is lost in a new compression cycle. For a photograph you may open, edit, restore, and resave multiple times, TIFF ensures there is no cumulative quality loss. TIFF files are substantially larger: a 600 DPI scan of a 4x6 print saved as uncompressed TIFF is approximately 40-50 MB, while the same scan as a JPEG at 90% quality is around 3-5 MB. For everyday use and sharing, export a JPEG copy from your TIFF archive — use the TIFF as the permanent record. For AI restoration purposes, JPEG at high quality (90% or above) is functionally equivalent to TIFF as input for tools like Real-ESRGAN and NAFNet, because the models are designed to work with the compression artifacts that JPEGs introduce. The most important variable is resolution, not format.

How should I organize scanned photos to make restoration easier?

A consistent naming and folder convention makes it much easier to track which photos have been scanned, restored, and finalized. A practical structure for a home digitization project: create a top-level folder named "Family Photos Digitization," with subfolders for each source collection ("Mom childhood photos," "Dad military service," "Grandparents wedding") and within each collection, subfolders for each stage: "raw-scans," "restored," and "final-prints." Name files with a date prefix and a descriptive label, such as "1958-summer-vacation-beach-01.tiff." Date-first naming ensures files sort chronologically in any file browser. Keep raw scans permanently — even after restoration, the unmodified scan is the closest thing you have to a digital original and should never be deleted or overwritten. After restoring through ArtImageHub or a similar tool, save the downloaded restored file in the "restored" subfolder with the same base name plus "-restored" added. This way, if AI restoration technology improves in future years, you can always run the raw scan through a newer tool and compare results without having lost the original.