Metadata
What photo catalog software is for
Photo catalog software doesn’t sort files into folders. It keeps a database that says, for each image, what it is, where it lives, what it shows, and what is known about it. Over the years, that database becomes nearly as valuable as the images themselves: it’s what lets you find a photo, prepare an exhibition, answer a gallery, or hand over a body of work.
So the choice commits you for a long time. Product pages all look alike; the differences show up in the details, often after months of data entry. The nine criteria below can be checked before you commit. For the basics, see What Is a Photo Catalog, and Why Isn’t It Just a Folder?
Before comparing: a simple method
- Build a test folder of a few hundred images: RAW files, JPEGs, files already described in another program, a few duplicates.
- Think ten years ahead, not just next week: what will you need to find, prove, or pass on?
- Read the documentation. For each criterion, look for the written answer. A feature that isn’t described anywhere doesn’t exist for you.
The nine criteria
1. Originals stay where they are
A catalog should reference your files, not move them into a folder only it controls. Ask how it recognizes a file, too: by its name, which is fragile, or by its content. A content fingerprint finds a duplicate even after it’s been renamed, and tells you when a file has changed.
2. Metadata stays readable elsewhere
EXIF, IPTC, and XMP are standards. The software should read them and write your descriptions in a form other tools can read back. Check where it writes: only in its own database, in an XMP sidecar next to the image, or in the original itself. Writing into an original should always be a deliberate choice. See What Is a Sidecar XMP File?
3. A controlled vocabulary, not just keywords
Free keywords drift: “Paris”, “paris”, “Paris FR”. A thesaurus arranges terms from broad to narrow and attaches synonyms to them. Check that search actually uses it, and that the vocabulary can be exported.
4. Finding a photo without knowing its file name
Search across titles and descriptions, filters by rating or date, hand-built collections, collections that update themselves, saved searches. Test with a real question: “portraits from 2019 rated three stars or more.”
5. The work is not the file
If you document your work as an artist, the software must tell the work, the unit you cite, apart from the files that show it and the prints that exist of it. Without that distinction, a scan, a master, and an installation view become three “works.” See How to Build a Photographic Catalogue Raisonné.
6. Confidentiality is built in
Prices, buyers’ names, living people: some data must not leave. Check that you can mark it confidential and that exports leave it out by default. Also check where image analysis, faces or similarity, takes place: on your machine, or somewhere else.
7. What you already have comes in undamaged
You probably already have a catalog in another program, or an inventory in a spreadsheet. The software should be able to take it over, show what it would do before writing anything, and leave the source untouched.
8. Data leaves in open formats
A catalog you can’t leave is a risk. Ask which export formats exist (spreadsheet, museum exchange formats), whether they hold all the data, and what you need to access them.
9. A clear place in the preservation chain
A catalog describes; it doesn’t guarantee that files are intact. Check how it backs up its own database, and how it fits with what comes before (gathering scattered drives) and after (preserving the originals).
How Obscura Catalog meets these criteria
Obscura Catalog is the Obscura Flow suite’s software for description. Here is what its guide establishes, criterion by criterion.
- Originals in place. Importing a folder copies and moves nothing: each image is referenced where it is, with its SHA-256 fingerprint, which recognizes a file already present even under another name. When a file changes, its new fingerprint becomes the reference and the old one is kept as a version. See Import a folder.
- Metadata. Catalog reads EXIF, IPTC, and XMP. By default it never writes into your files: your edits go into the catalog and into an XMP sidecar that other photo software reads back. Writing into the file itself is offered only for JPEG, TIFF, PNG, and HEIC/HEIF, and is confirmed on every save; for a RAW file, it’s always the sidecar. See Where to save.
- Thesaurus. Terms are arranged in a hierarchy, each with its synonyms. In search, a term also finds its synonyms and narrower terms. The thesaurus exports to CSV and SKOS. See The thesaurus.
- Finding. Quick search across names, titles, descriptions, keywords, creators, and indexing; advanced search by rating, label, tag, folder, and date; collections, smart collections, and saved searches. See Collections, tags and search.
- Work and file. A work is distinct from a file: it can have several, or none if it is lost or destroyed. Its catalogue number is never reassigned. Each record holds attributions with their degree of certainty, prints and editions, and provenance. See Work and file and Attributions, prints and provenance.
- Confidentiality. Confidential works and events are left out of every export; only the CSV, the full export, and the Vault package can include them, with a justification, for a single export. Internal notes never leave. A provenance event involving a living person stays confidential. Face and similarity analysis runs on the Mac: no image is sent anywhere. See The confidentiality rule and Privacy.
- Taking over what exists. Catalog imports a Lightroom Classic catalog read-only, after a dry run that writes nothing, and a CSV inventory, after a preview. See Import a Lightroom catalog and Importing a CSV inventory.
- Open formats. A full export of the library in open formats, CSV, LIDO 1.1, and CIDOC CRM. The full export, the Vault package, and CSV are included without a licence: your data always leaves in an open format. See Exporting the catalogue.
- Preservation chain. Catalog backs up its own library. Upstream, Obscura Trace inventories your drives, finds duplicates, and consolidates through verified copies, deleting nothing without your consent. Downstream, Catalog prepares a package for Obscura Vault, which archives it, checks its fixity, and keeps an offsite copy.
Traps to avoid
- Judging by the interface. A pleasant grid tells you nothing about what happens to your data in ten years.
- Entering data before exporting. Run a full export of your test folder on day one, and open it in a spreadsheet.
- Letting keywords multiply. Set up a first vocabulary before mass entry; fixing it afterwards costs far more.
- Mistaking the catalog for a backup. The catalog describes your originals; it doesn’t protect them. You need both.
- Migrating everything at once. Bring over a representative sample first and check the result before the rest.
Going further
If you want to describe your work all the way to a catalogue raisonné, start with Obscura Catalog and its Import a folder chapter: one test folder is enough to check all nine criteria on your own files. To see how the suite’s four programs share the work, from your drives to preservation, visit the software page or the documentation.