How to Digitize Old Family Photos (Without Never Finishing)

Most photo-digitisation projects fail the same way. Someone buys a scanner, starts with the oldest album, does a beautiful job on the first eleven pages, and stops. Two years later the scanner is in a cupboard and the box is exactly where it was.

The failure is not effort or equipment. It is sequencing.

The thing with a deadline is identification, not resolution#

A photograph has two components: the image, and everything the image does not contain. Who these people are. Where this is. What year. Why everyone is laughing. Which house that is. That this was the last summer before the move.

The image is stable. Paper deteriorates slowly and predictably, and a print from 1962 that has survived to now will very likely survive another decade in a shoebox.

The second component is not stable at all. It lives in a small and shrinking number of heads, and it disappears without any warning event. Nobody is present at the moment a face becomes permanently anonymous; it just happens, and afterwards the photograph is a picture of strangers who happen to be your family.

So the correct first move is not scanning. It is sitting down with the oldest person you have, and a stack of prints, and recording them saying who everyone is.

Twenty photographs annotated by an eighty-year-old are worth more than two thousand scanned at 1200 dpi and captioned by nobody.

Work in passes#

Pass one — fast and complete. Go through the entire box. Photograph each print with a phone in indirect daylight. Do not sort, do not crop, do not colour-correct, do not stop to admire anything. The goal is a complete rough index of what exists, and it should take an evening or two, not a season.

Photograph the back of anything with writing on it. That handwriting is frequently the only primary source for a date or a name, and it is the first thing that vanishes when a print is scanned face-only.

Pass two — identification. Take the rough index to the people who can read it. Record them. Do not try to type notes while they talk; you will miss the good part, which is always the digression.

Pass three — quality, for the few that deserve it. Now, and only now, decide which images justify a real scan: the portrait that will be enlarged, the wedding photograph, the one of the house. That is usually a few dozen images, not the whole box, and a scanning service is entirely reasonable for it.

Most people who try to do pass three first never reach pass two, and pass two is the one that matters.

Practical capture, with a phone#

Indirect daylight near a window. Never direct sun; never flash. Lay the print flat on a plain surface — a sheet of white paper works — and hold the phone parallel to it rather than tilted, or the rectangle will keystone and you will notice forever.

Block reflections with your own body rather than moving the print. If a photograph is behind glass in a frame, take it out; the glass will cost you more than the handling risk.

Shoot in the highest quality your phone offers, and turn off any "beautify" or heavy processing.

Practical capture, with a scanner#

600 dpi for standard prints is the sensible default and gives room to enlarge and reprint. Go higher — 1200 dpi and up — for anything small, and for negatives and slides, where you are magnifying significantly.

Save as TIFF or maximum-quality JPEG for the masters. Keep the masters untouched and do any cropping or correction on copies; a re-cropped, re-saved JPEG loses information permanently every time.

Handle originals by the edges, and do not use adhesive, tape, rubber bands or paperclips on anything you want to keep.

Slides, negatives and the awkward formats#

Slides and negatives frequently contain images that were never printed, which means a box of them can hold pictures no living person has seen. They are also the format most likely to be discarded during a house clearance because nobody can view them.

If you have them, prioritise a scanning service over a hobby setup — the per-item cost is low relative to how much unique content is typically in there, and the failure mode of a cheap slide attachment is a box you look at once.

Cine film and video tape are a separate problem with a harder deadline; magnetic tape degrades in a way paper does not.

What to do about the physical originals#

Digitising is not a reason to throw prints away, but it does change what "keeping" costs. Store what you keep flat, in a cool, dry, dark place, away from exterior walls, lofts and basements — the two enemies are heat cycling and humidity, and lofts supply both.

Split irreplaceable originals between two locations if you can. A single box in a single house is one flood away from being a story about a flood.

Storage is not preservation#

The most common end state of a successful digitisation project is nine thousand files in a folder called Scans, on one laptop, under one login, with no captions.

That is not an archive. It is the same problem with better lighting.

The minimum bar for calling this finished: the files exist in more than one place; someone other than you can find and open them without your password; and the identification you collected is attached to the images rather than living in a separate recording nobody will ever line up again.

That last part is the one people skip, and it is the reason the whole exercise was worth doing.