Fixity Checks at Home: Keeping Fingerprints of Your Files
Learn how to record and compare file checksums so you can spot silent changes in your archive before they spread.
Archivists have a word for confirming that a file has not changed: fixity. It sounds technical, but the idea is very down to earth. You take a kind of fingerprint of a file when it is in good condition, and later you check whether the fingerprint still matches. If it does not, something has gone wrong, and you can restore the file from another copy before the damage spreads. This practice is the best answer to the silent damage described in our article on bit rot and why digital files degrade.
What You Are Actually Recording
A checksum is a short string of letters and numbers calculated from the exact contents of a file. Change a single bit, and the string changes. Common types include MD5 and SHA-256. For modern archives, SHA-256 is a good choice since it is widely supported and considered robust.
When to Record Checksums
The best moment is right after a collection is complete and you have verified that all files open correctly. Recording fingerprints of damaged files only preserves the damage. For collections that grow, such as an ongoing photo library, record values for each batch as you add it.
Step 1: Generate the Values
There are several ways to do this, depending on your comfort level:
- Command-line tools are built into most systems. On Windows, PowerShell offers a hash command; on Mac and Linux, tools such as shasum and sha256sum do the job.
- Graphical tools let you select a folder and produce a list of checksums with a click. Choose one from a reputable source and from the developer's own site.
- Backup and synchronization programs sometimes include verification features that compare files without a separate step.
Whichever you choose, test it first on a small folder of copies so you understand the output.
Step 2: Save the List Alongside the Collection
Store the results in a plain text file within the collection folder, with a name such as checksums-2025-06. Each line normally shows the checksum followed by the file name. Keep another copy of this list somewhere separate from the files, so that a problem in one place does not take both.
Step 3: Check on a Schedule
Pick a rhythm you can maintain, perhaps once or twice a year. Recalculate the checksums and compare them with the saved list. Many tools can do this automatically, flagging any file that differs or is missing. Pair the check with your regular review of backups and your inventory of collections.
What to Do When a Mismatch Appears
- Do not panic, and do not overwrite anything.
- Check the same file in your other copies. If a copy still matches the saved fingerprint, it is good.
- Replace the damaged file with the healthy one.
- Look at the device that held the bad copy. Repeated problems suggest failing storage that you should replace.
- Record in your notes what happened and the date.
A mismatch is not always corruption. If you edited a file on purpose, its checksum will legitimately change. That is why finished archive material should be kept separate from working files.
Keep It Proportionate
You do not need checksums for everything. Focus on irreplaceable collections: family photos, scanned records, creative work and home videos. For everyday documents that you can easily recreate, normal backups are enough.
A Simple Example Routine
After finishing the scan of a stack of old letters, you open each file, confirm it looks right, generate checksums and save the list. You copy the folder to a backup drive and compare the checksums there as well. Twice a year you rerun the check on both copies. The whole process takes minutes and gives lasting peace of mind.
Frequently Asked Questions
Is a checksum the same as a backup? No. A checksum only detects change. It cannot repair a file. You still need separate copies to restore from.
Which checksum type should I use? SHA-256 is a good default. MD5 can detect accidental errors but is weaker against deliberate tampering.
Will checking checksums wear out my drives? Reading files has little effect. Annual or twice-yearly checks are very light use.
Steps
- Generate checksums: Use a checksum tool or a built-in command to create a fingerprint value for each file in a finished collection.
- Store the list safely: Save the list of values in a text file inside the collection and keep a copy in a separate place.
- Compare on a schedule: Every six to twelve months, recalculate the values and compare them with the saved list to find any changes.