CousinChart โ€” family connections made simple

How to Digitize Family Documents

Digitising is the cheapest insurance in genealogy: it protects irreplaceable papers against fire, flood and the relative who clears a garage without asking, and it makes a shoebox searchable. The parts that matter are scanning at a high enough resolution to be useful in twenty years, saving in formats that will still open, and having a backup that isn't a single hard drive. Everything else is refinement. This page covers settings by document type, a naming system that works, OCR, and a realistic backup plan. Keeping the people straight while you sort papers is easier with the printable family tree alongside.

Guide to scanning and digitizing family documents
๐Ÿ’พ What are you scanning? โ€” tap one:
Pick one ๐Ÿ‘‡
Get the printable family tree โ†’

What Settings Should You Use?

ItemResolutionColorMaster format
Certificates, letters, papers400โ€“600 dpiColor, even if the item looks grayTIFF
Photographs600 dpiColorTIFF
Small photos, under 4ร—5 in1200 dpiColorTIFF
Negatives and slides2400โ€“4800 dpiColorTIFF
Newspaper cuttings400โ€“600 dpiColorTIFF
Multi-page documents300โ€“400 dpiColorPDF + TIFF pages
Oversized itemsPhotograph, or section-scanColorTIFF

Always scan in color, even for a black-and-white document. Color captures paper tone, foxing, ink variation and faint pencil that greyscale flattens away โ€” and you can convert to greyscale later, but you can't get the information back.

Don't use "auto" enhancements. Automatic sharpening, dust removal, contrast and color correction all discard data. Scan flat and neutral, then edit a copy.

Scan at the highest resolution you'll realistically finish. A 600 dpi scan you actually complete beats an 1800 dpi one abandoned at item forty.

Include a ruler or scale in shots of unusual items, so future viewers know the size.

Scan the reverse of anything with writing, stamps, or a photographer's imprint on it.

What File Formats Actually Last?

The question people get wrong, and it matters in twenty years rather than today.

TIFF for masters. Uncompressed or losslessly compressed, universally supported, no generational loss. Large files โ€” 20 to 100 MB each โ€” and storage is cheap.

JPEG for sharing and everyday use. Small and convenient, and it loses quality every time it's re-saved. Fine as a derivative, wrong as a master.

PDF for multi-page documents โ€” a will, a letter of several pages, a service record. PDF/A is the archival variant if your software offers it.

PNG is fine for screenshots and lossless web images, less standard for archival scanning.

Avoid: proprietary scanner formats, RAW-only workflows without a TIFF export, and anything that requires one specific piece of software to open.

The rule of thumb: keep one unedited archival master per item, plus whatever derivatives you need. Never edit the master, never re-save a JPEG repeatedly, and never delete an original after scanning it.

How Should You Name and Organize Files?

The part that decides whether a collection is usable or just large.

A naming convention that works:

`YYYY-MM-DD_surname-firstname_document-type_place.tiff`

For example: `1901-04-12_hutchings-mary_birth-certificate_blackburn.tiff`

Principles:

Folder structure, two workable approaches:

Either works; mixing them halfway through does not.

Keep a spreadsheet index listing filename, person, date, place, document type, where the original is stored, and where you obtained it. It's also what makes citation straightforward later, since everything you need is already in one row.

Record where the physical original lives. "Original with Aunt Jean, Sheffield" is information that vanishes otherwise.

What About OCR and Transcription?

Scanning captures an image; OCR and transcription make it searchable, which is where the real value is.

OCR works well on printed text, typed documents, modern newspaper print, and typed service records.

OCR works badly on handwriting, faded ink, old typefaces, and anything with heavy foxing. Handwritten records still generally need a human, though handwriting recognition is improving quickly.

A practical approach: run OCR where it's likely to work, and hand-transcribe the important handwritten items โ€” wills, letters, diaries. Transcribe exactly, including spelling and abbreviations, then add a note explaining anything unclear rather than silently correcting it.

Mark uncertainty in transcriptions with square brackets: `[illegible]`, `[Elizabeth?]`. Silently guessing turns a transcription into a source of future errors.

Save transcriptions as plain text or a simple document alongside the image, with a matching filename. Plain text will open in fifty years; a proprietary format may not.

Why bother: a transcribed collection is searchable. Typing a surname and getting every mention across two hundred documents is a different research experience from remembering which shoebox something was in โ€” and quoted transcriptions are what bring a narrative to life in how to write a family history.

Where Do You Find the Documents in the First Place?

Worth a section, because most people underestimate what's still sitting in family houses.

Ask specifically, not generally. "Do you have any old papers?" gets a no. "Do you still have Grandad's army discharge papers, or the deeds from the old house?" gets a search. People don't think of certificates, indentures, ration books, employment cards, insurance policies and school reports as family history until you name them.

The places things turn up: the back of wardrobes, inside books, in the linings of suitcases, behind photographs in frames, in the pockets of stored clothes, and in biscuit tins. Family bibles frequently have loose papers tucked in.

Ask at the same time as an interview. Papers prompt memory and memory prompts papers, so combining the two is far more productive than doing either alone โ€” see how to interview a relative.

Offer to scan and return. Relatives are far more willing to hand something over for a week than to give it away permanently, and a scan is all you need. Return it promptly and with a printed copy for them, and you'll be offered more next time.

Scan the boring things too. Rent books, receipts, wage slips and ration cards carry addresses, employers, dates and signatures โ€” often the detail that breaks an impasse later.

Don't forget your own generation. Certificates, letters and photographs from the last thirty years are family history too, and they're the ones nobody thinks to preserve because they don't feel old yet.

What Backup Plan Actually Protects You?

The 3-2-1 rule, borrowed from professional practice and worth following exactly:

In practice that might be: the working copy on your computer, a copy on an external hard drive kept in the house, and a cloud backup or a drive at a relative's house.

Why one drive isn't enough: hard drives fail without warning, cloud accounts get closed, houses flood, and laptops get stolen. Every one of these regularly destroys years of genealogy work, and the loss is total because the scans were the only copy.

Practical points:

Leave instructions. A note explaining where the files are, how they're organized, and what the passwords situation is means the collection survives you. A great deal of digitised family history is lost simply because nobody else knew it existed.

The physical originals still matter after scanning โ€” store them properly rather than discarding them, as covered in how to preserve old family photos. And once the collection is scanned and labeled, it becomes the raw material for something people actually read, which is the point of how to create a family history book.

What Equipment Do You Actually Need?

Less than the marketing suggests, and the cheap end is genuinely adequate.

A flatbed scanner is the core purchase, and a ยฃ60โ€“ยฃ120 model handles documents and photographs perfectly well at 600 dpi. Where it's worth spending more is a transparency unit for negatives and slides, which cheaper models don't include.

A phone camera is a completely reasonable substitute for anything flat that won't fit a scanner, and the only sensible option for bound books and fragile items. Modern phone cameras out-resolve what most people scan at.

A small tripod or phone stand matters more than the camera. Handheld shots are never quite square, and the distortion is obvious when you come to read a document.

A window with indirect daylight beats any lamp you'll buy. Direct sun causes glare and hard shadows; a bright overcast day is close to ideal. Never use flash.

A sheet of plain gray or black card as a background makes edges clear and stops light bouncing through thin paper.

A piece of clean glass to hold a curled document flat while photographing โ€” but only on items sturdy enough to take it.

What you don't need: a document camera, a dedicated film scanner unless you have hundreds of negatives, expensive software, or a scanner with a sheet feeder โ€” feeders damage old paper and should never be used on anything irreplaceable.

Sheet-feed scanning services are worth considering for large volumes of modern, robust material. Never send anything fragile, irreplaceable or one-of-a-kind through one.

Frequently Asked Questions

What resolution should I scan family documents at?

400โ€“600 dpi for certificates, letters and papers; 600 dpi for photographs, or 1200 for small ones; and 2400โ€“4800 dpi for negatives and slides. Always scan in color, even for black-and-white items, since color captures paper tone and faint marks.

What file format should I use for archival scans?

TIFF for the archival master โ€” uncompressed, universally supported, no quality loss on re-saving. JPEG for sharing, and PDF for multi-page documents. Keep one unedited master per item and edit copies rather than the original file.

Should I scan or photograph old documents?

Scan flat items where the scanner lid can close gently. Photograph bound books, fragile or torn papers, and anything oversized โ€” a scanner lid presses and a spine forced flat can break. Use indirect light, no flash, and hold the camera square.

How should I name scanned genealogy files?

Date first in YYYY-MM-DD form, then surname-firstname, then document type and place โ€” for example `1901-04-12_hutchings-mary_birth-certificate_blackburn.tiff`. Use lower case with hyphens, no spaces, and apply it consistently.

How many backups do I need?

Three copies, on two different types of storage, with one off-site. A single external hard drive is not a backup. Automate the cloud copy, test a restore occasionally, and give copies to relatives in other households.

Can I throw away originals after scanning them?

No. Scans are insurance, not replacements โ€” originals carry paper, ink, watermarks and physical evidence a scan can't capture, and future technology may extract more from them. Store the originals properly instead.


A final word on pace. Digitising a family collection is not a weekend job, and treating it as one is how it gets abandoned. Twenty items a session, properly named and backed up, will clear a substantial archive over a year โ€” and a partially digitised collection with a good naming system is far more useful than a fully scanned one dumped into a folder called 'scans'.

Scan in color at a resolution you'll actually finish, keep unedited TIFF masters, name files so they sort themselves, and get three copies with one off-site. Do that and a shoebox becomes a searchable archive that outlives the paper โ€” and outlives you. Keep the relationships straight as you work through it with the printable family tree.