Preservation
“Archive” means several different jobs
Search for photo archive software for Mac and you’ll find tools that have little in common: photo editors with a catalog, digital asset managers, backup utilities, cloud services, archival description tools. Each one does part of the job. None of them, on its own name, tells you whether your originals will still be readable and intact in twenty years.
For a photographer or an artist, archiving means at least four things: knowing what you own, describing it, keeping verified copies of it, and being able to get it back. The criteria below apply whatever tool you choose. Check them during a trial, on a copy of a real folder, not on the vendor’s demo.
Nine criteria for photo archiving software
1. It leaves your originals alone
An archive tool should read your files, not rewrite them. Ask where your edits to titles, keywords and captions go: into the file itself, into a separate catalog, or into an XMP sidecar next to the file? Writing into the original changes its fingerprint and, for some formats, means re-encoding it. That should be an explicit choice, never a default. Our article on writing metadata into originals or into sidecars covers the trade-offs.
2. It can prove a file is intact
A file can degrade without any warning. The only way to know is to read it in full and compare its fingerprint (a SHA-256 checksum, for example) with the one taken when it entered the archive. Look for scheduled checks, a history that is never rewritten, and a clear difference between “verified” and “couldn’t be checked”. A disconnected drive isn’t good news. See what fixity means.
3. It finds duplicates and lets you decide
Exact duplicates (identical content) and visual duplicates (a resized or re-saved copy) aren’t the same thing, and shouldn’t be treated with the same confidence. A good tool explains why it groups files and recommends which to keep, but nothing should be deleted without your say.
4. It speaks standard metadata
EXIF, IPTC and XMP are the shared languages of photo metadata. The tool should read them and write them in a form other software can read. Otherwise your descriptions are locked inside one application.
5. It helps you find images again
Search across file names, titles, keywords and dates is the minimum. For a large body of work, a controlled vocabulary with synonyms and hierarchy (so that “birds” also finds “gull”) keeps descriptions consistent over the years.
6. It keeps several copies, and checks them
Three copies, on two different media, one of them off-site: the 3-2-1 rule. A serious tool knows where each copy lives, verifies a copy before sending it elsewhere, and reminds you when a drive is getting old.
7. It restores, and checks the restore
A backup only counts once it has been restored. Look for a restore that verifies the archive before extracting anything, never overwrites an existing file, and checks each restored file against the inventory.
8. It lets you leave
Any software will one day be replaced. Ask what is left without it: are the archives in open formats, readable with standard tools? Can the catalog be exported in full, in formats an institution can read?
9. It works locally
Your images and their descriptions should stay on your Mac unless you decide otherwise. Off-site copies are necessary, but you should choose where they go. More on this in local-first archiving.
For professionals, one more question matters: does the tool describe the work, not only the files? A catalogue raisonné distinguishes a work from its prints and files. A studio’s papers call for archival description.
How the Obscura Flow suite answers, app by app
Obscura Flow is a suite of four macOS apps. Each one covers a part of the job, and they exchange files, not a shared database.
Obscura Trace: knowing what you own
Obscura Trace inventories your drives and finds exact duplicates (identical SHA-256) and visual duplicates. In the Review Center, it recommends a master and explains why, criterion by criterion; you decide, and no decision modifies a file. The Transfer Center then consolidates what you keep: copy, fingerprint check, and only then removal of the source, which goes to the Trash. A JSON manifest records where each file came from. Trace works entirely on your Mac: no analysis engine uses the network.
Criteria met: 1, 3, 9.
Obscura Catalog: describing the work
Obscura Catalog references your images where they are, with their SHA-256 fingerprint, and reads EXIF, IPTC and XMP. By default it never writes into your files: edits go to the catalog and an XMP sidecar; writing into the file is offered for some formats only, and confirmed each time. A thesaurus with hierarchy and synonyms feeds the search, and exports to SKOS. Face and similarity analysis runs on the Mac; no image is sent elsewhere. Catalog also keeps the catalogue raisonné (works, attributions, prints, provenance) and exports the full library in open formats, plus LIDO 1.1, CIDOC CRM and CSV.
Criteria met: 1, 4, 5, 8, 9, and the description of the work.
Obscura Vault: preserving and proving
Obscura Vault records each file’s fingerprint at import and never replaces it. Fixity audits, on demand, weekly or monthly, re-read every file. Each file is Verified, Altered, Not verifiable or Never checked; a file never checked is never counted as verified. Projects are packed into signed .obf archives, optionally encrypted. The storage media inventory maps the 3-2-1 rule to three levels and gives each drive a replacement date. Offsite copies go to S3 or compatible storage, a remote folder or a synced folder, after the local archive has been verified. A restore verifies the archive first, never overwrites, and reads back every file. An unencrypted .obf archive stays readable without Vault, with standard tools.
Criteria met: 2, 6, 7, 8.
Memory Atlas: arranging the studio’s archives
Memory Atlas describes archives (letters, contracts, scans) to the ISAD(G) standard, from fonds down to item. Files stay in place: Atlas links to them without copying them. Descriptions export to EAD 2002 or EAD3, checked against the official schema before being saved, and as a printable finding aid.
Criteria met: 1, 8, and archival description.
The criteria at a glance
| Criterion | Where to look in the suite |
|---|---|
| Originals left alone | Trace (Trash, verified copy), Catalog (XMP sidecar by default), Atlas (files linked, not copied) |
| Proof of integrity | Vault: fingerprint at import, fixity audits, four verdicts |
| Duplicates under your control | Trace: exact and visual, recommended master, your decision |
| Standard metadata | Catalog: EXIF, IPTC, XMP |
| Finding images again | Catalog: search, smart collections, thesaurus |
| Several verified copies | Vault: 3-2-1 levels, media inventory, offsite copies |
| Verified restore | Vault: verify, extract without overwriting, read back |
| A way out | Catalog full export, Vault .obf format, Atlas EAD |
| Local by default | Trace and Catalog analyses run on the Mac |
One app or four?
You don’t need all four. Catalog and Vault, for example, share no database and no settings: Catalog writes a package for Vault, Vault verifies and archives it, and can give it back byte for byte. Each app can be used on its own, and the files they exchange remain readable.
If you already know your files and want proof that they’ll last, start with Obscura Vault. The suite’s software page shows where each app fits, and the documentation gathers all four guides.