Sealion Projects

Dig Site 3026

A deadpan site report from 3026 on seven places people keep their photos, and the year a dig stops finding each one.

dataviz · desktop-firstData as of 1 Nov 2026Published 1 Nov 2026archivephotosdigital-preservationreport

Interactive explainer

Sources and method

The question

Most people keep their photos in two or three places and assume at least one of them is permanent. Each place has a published lifetime, but the figures sit in very different documents: a cloud provider's account policy, a federal standards report, a printer-ink lab rating, a photo-paper data sheet. Nobody lines them up. This piece puts seven common ways of keeping photos on one timeline running from 2026 to 3026, and asks a deadpan archaeologist what a dig would find at four points along it.

The data

Each deposit carries a published range in years: the lower figure is what its source says to expect, the upper is the most the sources support. A cloud account can be deleted after 2 years without a sign-in (Google, 2023). NIST IR 8387 (2022) gives unpowered SSDs under a year, burned CDs and DVDs a conservative 30 years with reputable claims of 50 or 100, and M-DISC at least 100. Wilhelm's 2004 display ratings put lab prints framed under glass at 14 to 40 years. Fujifilm's album paper data sheet gives 10 to 20 years of almost no change in the dark below 25°C, and Wilhelm's earlier work quotes 60 years in dark storage. The 10-year ceiling for phones and laptops is our reading, marked as editorial on the page.

The form

A section drawing, the way a dig report draws its trench: one row per deposit, a log-scale time axis so that one year and one thousand years both stay readable, and four dig columns at 2027, 2036, 2126 and 3026. A solid strip runs to the expected figure, a hatched strip to the upper one, and a stamp in each column records the find as intact, fragmentary or absent. The finds register below repeats every row as text, in the order the finds are lost, for anyone not reading the drawing.

What it leaves out

These are rated and specified lifetimes, not observations of anyone's actual photos. A print in a humid Singapore flat will not behave like one at 25°C, and a disc that is scratched or kept in a hot car will not reach its rating. The display ratings measure fading to a set endpoint, not the paper falling apart, so "absent" for a print means the picture is gone, not the object. Ratings for inkjet prints range far wider than for lab prints, and newer papers may rate higher than the 2004 figures. Diaries, film negatives and tattoos are left out because we could not cite a survival figure for them.

Sources

All retrieved 2026-11-01.

A focused prototype from Sealion Projects. Anything it saves stays in this browser, on this device.