Genealogy Photo Scanning: Capture Once, Archive Forever

The shoebox arrives in the mail from your aunt: 412 family photos spanning 1920 to 1980, the corners curled, the prints starting to fox at the edges where humidity got in. Your grandmother is the last person alive who can identify some of these people, and she's 91. You sit down with the scanner you bought for your home office, set the settings to "fast preview" because there are so many photos to get through, and start digitizing at 200 DPI as JPG. Fifteen years later your kids find them on an old hard drive in a folder called "Photos." Nobody knows who anyone is, the scans aren't sharp enough to crop a tight portrait, and the JPG quality is too low to print a memorial photo when your father passes.

Family history photo work is unforgiving: you usually only get one chance to scan an original before it gets lost, fades, or burns in a house fire. The settings you pick now determine the quality ceiling for everything your descendants can do later. The metadata you embed now determines whether your great-grandchildren can identify the people in the photos or whether they're staring at strangers. The naming convention you adopt now determines whether the archive remains organized across 50 years of hardware changes. This guide is the multigenerational specification for doing genealogy photo work properly the first time.

Scanner settings: capture once, capture right

Scan every original at 600 DPI minimum for prints up to 5x7, and 1200 DPI for anything smaller than 4x6 or with fine detail you might want to crop into later. Use the scanner's TIFF output mode for the archive master, not JPG. A 4x6 print scanned at 1200 DPI as a TIFF is roughly 100 MB; that's the file you preserve for generations. Then derive a JPG copy for sharing.

Background: why amateur scans fail their descendants

The dominant amateur scanning pattern from roughly 2000 to 2015 was 200 DPI JPG at quality 70, saved to whatever folder structure made sense at the time. The resulting files are too small to crop, too compressed to enlarge, and named in patterns that broke the moment the family member who created them died. Modern genealogy practice has shifted toward archival-grade capture (600+ DPI TIFF masters), strict naming conventions, embedded metadata, and the 3-2-1 backup rule. The shift took twenty years and a lot of lost archives to learn.

The TIFF master plus JPG derivative split

Every photo you scan should produce two files: a TIFF master and a JPG derivative. The TIFF lives in your archive (external drive, cloud, NAS, all three ideally), is never edited, and never gets shared. The JPG derivative is what you email to cousins, post to family trees on Ancestry, and print for reunions. When you need a sharper version five years from now for a memorial slideshow, you re-derive from the TIFF instead of upscaling a stale JPG. Use TIFF to JPG to batch-create derivative copies from your archive at any size.

Naming conventions tied to surname and year

Adopt a strict naming pattern before you scan a single photo: SURNAME_YYYY-MM-DD_PersonName_SequenceNumber.tif. For example: HARRIS_1947-06-12_RuthAnnHarris_001.tif. If the date is approximate, use YYYY-00-00 or YYYY-MM-00. If the person is unidentified, use UnknownMale or UnknownFemale with the apparent decade. The naming makes the archive sortable by surname, by year, and by sequence, which is critical when you have 4000 files.

Dust, scratches, and color correction

Modern flatbed scanners like the Epson V600 and V850 include ICE (Infrared Cleaning Element) that detects dust and scratches during the scan and removes them automatically. Turn ICE on for color photos; turn it off for true black-and-white silver prints (it can soften them). Do color correction on the TIFF in 16-bit mode before deriving the JPG; baking corrections into the JPG is fine for the share copy, but the TIFF master stays unedited.

Handling damaged originals

Torn corners, water stains, faded color, foxing on the paper: scan the photo as-is, then make any repair work in a derivative file. Never restore on top of the master. The principle is that a future generation might have better restoration tools, and they need the unrepaired scan to start from.

Building shareable family albums

Once you have JPG derivatives, group them into family albums and export as PDF for distribution. A 50-photo album wrapped in a single PDF is easier to email or share via Google Drive than 50 separate JPGs. Use JPG to PDF to compile decade-by-decade albums (1920s, 1930s, etc.) with caption pages between groups.

Compress for sharing without touching the master

Your TIFF masters never get compressed. Your JPG derivatives at 2400 pixels on the long edge land around 1 to 2 MB each, which is fine for email and family tree uploads. If you're posting to a low-bandwidth platform (Ancestry, FamilySearch, MyHeritage), compress the JPG to about 500 KB using JPG compressor. The compression only affects the derivative, never the master.

Verifying scan metadata

Embed the date, location, and people in IPTC metadata fields of the JPG derivative so the information travels with the file. When you upload to a genealogy site, the platform often reads IPTC and pre-fills caption and date fields. Verify metadata stuck with image info by uploading the JPG and reading back the embedded tags.

Storage strategy: the 3-2-1 rule

Keep 3 copies of every photo, on 2 different media types, with 1 copy off-site. For a 200 GB family archive that's: copy 1 on your main computer's internal drive, copy 2 on an external USB-C drive in a drawer, copy 3 in a cloud archive (Backblaze, iDrive, Glacier). Verify the cloud copy every 6 months by downloading 10 random files and checking they open.

Format compatibility for ancestry sites

Ancestry.com, FamilySearch, and MyHeritage all accept JPG up to 15 MB per file. Some specialty European archives still require TIFF for high-resolution uploads. If you're contributing to a community-driven family tree, post the JPG derivative; keep your TIFF master private. Run your JPG derivative through the image converter if a specific site requires PNG (rare) or a particular pixel dimension.

Captioning workflow with relatives

The hardest part of genealogy photo work isn't scanning, it's identifying who's in each picture. Set up a shared Google Drive or Dropbox folder, invite older relatives, and ask them to leave comments on photos identifying people, places, and dates. Update IPTC captions in the JPG derivatives as identifications come in. Older relatives die, and the information they hold dies with them; the captioning project has to start while they're still around. A weekly Sunday call with grandparents going through 20 photos at a time can capture 1000 identifications in a year.

Step-by-step: scanning a single photo properly

1. Clean the scanner glass. Microfiber cloth, no Windex. Dust on the glass becomes a permanent feature of the master.
2. Handle the original by edges only. Cotton gloves help; fingerprints on prints accelerate fading.
3. Position the photo on the platen. Straight, no tilt. Use the scanner's edge guides.
4. Set scan parameters. 600 to 1200 DPI, TIFF output, 16-bit color depth, ICE on for color (off for B&W silver prints).
5. Scan and save to archive folder. Use the SURNAME_YYYY-MM-DD_Person_Sequence.tif naming convention.
6. Derive the JPG. Run the TIFF through TIFF to JPG at 2400 pixels long edge, quality 90.
7. Embed IPTC metadata. Date, location, people, source (who owned this photo), copyright (likely "Public Domain - family use").
8. Verify with image info. Confirm IPTC fields stuck and color space is sRGB.
9. Back up immediately. 3-2-1 rule applies the moment the file exists, not at end of session.

Common mistakes and how to fix them

• Mistake: scanning at 200 DPI to save time. Fix: rescan at 600 DPI minimum. The hour you save now costs your descendants the ability to crop and enlarge.
• Mistake: JPG-only archive with no TIFF master. Fix: rescan critical photos to TIFF. JPG re-edits compound compression artifacts.
• Mistake: no naming convention. Fix: adopt SURNAME_YYYY-MM-DD_Person_Sequence before scanning the next photo. Rename old files in batch.
• Mistake: editing the master TIFF. Fix: always edit derivatives, never the master. Future restoration tools will be better than today's.
• Mistake: single backup copy. Fix: implement 3-2-1 (3 copies, 2 media, 1 off-site). House fires destroy single-copy archives.
• Mistake: skipping IPTC captions. Fix: caption each photo in IPTC as identifications come in from relatives. The file carries the story.

Real-world examples

The Smithsonian's family history workshops teach the 600 DPI TIFF master plus JPG derivative pattern as the standard for hobbyist genealogists. Their downloadable spec sheet matches what's described here. FamilySearch's volunteer indexing program processes millions of public-domain photos at 400 DPI grayscale (or 600 DPI color) as TIFF masters, with JPG derivatives served through their public portal. MyHeritage's photo enhancer service requires 1200x1200 minimum upload and works best with TIFF or high-quality JPG sourced from 600+ DPI scans; lower-quality sources produce unconvincing AI restoration.

Print-type scan settings comparison

Print type	DPI	Format	Bit depth	ICE
Color print (post-1970)	600	TIFF	16-bit	On
Color print (pre-1970, faded)	1200	TIFF	16-bit	On
B&W silver print	600-1200	TIFF	16-bit grayscale	Off
Tintype, ambrotype, daguerreotype	1200+	TIFF	16-bit	Off
35mm slide	2400-4000	TIFF (transparency)	16-bit	On (color)
Polaroid	600	TIFF	16-bit	Off
Newspaper clipping	400	TIFF	8-bit grayscale	Off (descreen on)

Advanced tips for serious archivists

• Use a calibrated scanner profile. Generate an IT8 target profile for your specific scanner for accurate color reproduction.
• Scan both sides of every photo. Handwritten captions on the back are often the only ID you'll find.
• Build a relatives-attribution database. Track who provided which photo and which identifications, so future researchers can verify.
• Use OCR on newspaper clippings. Generate searchable text alongside the scan; it makes the archive queryable.
• Schedule annual archive audits. Verify drive health, refresh backups, spot-check 50 random files for corruption.
• Donate copies to local historical societies. Most county-level historical societies welcome digital donations of regional family archives.
• Plan inheritance. Write a one-page document explaining what the archive is, where it lives, how to access it. Store with your will.

FAQ

What DPI should I use for old slides versus prints?

Slides need 2400 to 4000 DPI because the original is small (35mm) and you want to enlarge. Prints need 600 DPI for 4x6 or larger, 1200 DPI for anything smaller or with fine detail.

Should I scan as TIFF or JPG?

Always TIFF for archive masters. Derive JPG copies with TIFF to JPG for sharing and uploading. JPG saves compound compression each time you re-edit.

How do I handle photos in old albums with adhesive backing?

Do not pull them off if they're stuck. Scan with the album page on the platen, then digitally crop. Adhesive removal often destroys the print surface.

What's the best scanner for genealogy work under $500?

Epson V600 covers most needs: prints, slides, negatives, decent ICE. The V850 is the upgrade for serious volume work.

Can I use phone-based scanning apps?

For casual sharing, yes. For archive masters, no. Phone scans top out around 300 DPI effective resolution and lack consistent color management.

How do I caption photos when nobody remembers who's in them?

Tag them "Unknown" with apparent decade and any context (estimated location, occasion). Future relatives may identify them; leave space for additions in the IPTC caption.

Should I share photos on Ancestry or keep them private?

Share JPG derivatives; never share TIFF masters. Public family trees benefit the community; the masters stay in your archive for quality control.

Color-corrected versus archival-faithful capture

Two valid philosophies compete in family photo preservation. Archival-faithful captures preserve the photo exactly as it exists today, including age-related fading and color shifts; this approach treats the original as historical artifact. Color-corrected captures restore what the photo presumably looked like when freshly developed; this approach treats the original as a window into the captured moment. The mature compromise is to maintain both: an unedited TIFF master (archival-faithful) and a color-corrected JPG derivative (presentation copy). Document both in IPTC so future viewers understand the difference.

Scanning oversized prints and panoramas

Prints larger than your scanner's bed (panoramas, 8x10 portrait prints on letter-bed scanners) require stitching. Scan the print in 30-percent-overlapping sections, then stitch in Photoshop's Photomerge or PTGui. The stitch quality depends on consistent lighting (don't open the scanner lid between scans) and accurate positioning. For prints up to 11x14, a tabloid-bed scanner (Epson V850 or Canon CanoScan 9000F) avoids the stitch step entirely. For prints over 16x20, camera-based copy stand setup with a 100mm macro lens beats scanner stitching on quality.

Photo album scanning without disassembly

Many family albums use magnetic-page or pocket-page construction where photos can be carefully removed for scanning. Old albums with permanent adhesive (1960s and 1970s photo corners or full-glue books) require scanning the whole page on a flatbed without removing the print. Then digitally crop each photo in post. The whole-page scan also preserves any handwritten captions on the album pages, which provide irreplaceable context.

AI-enhanced restoration: where it helps and where it doesn't

Tools like MyHeritage's photo enhancer, Remini, and Adobe's Photo Restore Neural Filter can dramatically improve faded or low-resolution scans. The catch: AI restoration hallucinates details that may not actually have been in the original photo, especially in faces. Use AI tools on derivative JPGs for sharing and printing, never on the TIFF master. Document any AI-enhanced version in IPTC as "AI-restored derivative" so future researchers know it's not a faithful capture. The restoration tools are improving annually; in 10 years they will be better than today's, which is another reason to preserve the original scan unmodified.

DNA-driven photo discovery

Modern genealogy has shifted toward DNA testing services (AncestryDNA, 23andMe, MyHeritage DNA) that connect users to distant relatives who may hold complementary family photos. Build a structured outreach process: when a DNA match emerges in your tree, ask if they have photos from a specific shared branch. Trade JPG derivatives (never originals) via Google Drive folder, with proper IPTC captions tagging the shared ancestors. The reciprocal exchange grows your archive while preserving each family's master ownership of their physical originals.

Long-term format strategy

TIFF as an archive format has been stable since 1986 and shows no signs of obsolescence. JPG has been stable since 1992. Both are likely safe for the next 50+ years. Avoid niche formats (proprietary scanner formats, vendor-locked archives, RAW formats tied to single camera brands) for masters because they may become unreadable as software vendors die. Plan a format-audit every 10 years: open a random sample of files, confirm they still open in current software, migrate if needed. Use image converter for batch migrations when format strategy changes.

Preserve once, share forever

Family photo archives are a multigenerational project, and the work you do now sets the quality ceiling for everything your descendants can do later. Scan to TIFF, derive shareable JPGs with TIFF to JPG, compress for online platforms using JPG compressor, bundle decade albums into shareable PDFs with JPG to PDF, swap formats in batch with image converter, and verify embedded captions through image info. Your grandchildren will thank you when they want to crop a portrait from a 1934 wedding photo and the master still holds detail.