Obscura Flow
Menu
Contact

Documentation

What Is Dublin Core for Photo Collections?

Fifteen simple elements, designed to be understood by any system. What Dublin Core brings to a photo collection, and where it stops.

IndexSupport

Metadata

A minimal foundation, deliberately thin

Dublin Core is a set of fifteen description elements — title, creator, subject, description, publisher, contributor, date, type, format, identifier, source, language, relation, coverage, rights — standardised as ISO 15836.

Its thinness is a choice. It never aimed to describe an object in detail; it aimed to let any system understand the essentials of a resource without knowing the domain. It is the smallest common denominator of documentary description.

What it brings to a photo collection

  • Immediate interoperability. A harvester, a portal, an institutional repository can read Dublin Core with no prior negotiation.
  • A starting discipline. The fifteen elements make a good checklist: if you can’t fill in creator, date, and rights, your description isn’t ready.
  • A safe exit point. Even if your internal model is rich, being able to project it into Dublin Core guarantees a third party can always get something out of it.

Its limits, worth knowing before committing

Dublin Core cannot express what makes a photographic collection specific: the degree of certainty of an attribution, the distinction between a work and a print, an edition of 3/8, a chain of provenance, a printing process.

That isn’t a flaw, it’s its scope. So a heritage collection needs a richer internal model — LIDO 1.1 or CIDOC CRM for museum exchange — and uses Dublin Core as a surface projection, not as a working model.

The common mistake

Filling in Dublin Core by hand, as a separate form. That’s work wasted twice: entry is redundant, and the two sets diverge at the first omission.

Dublin Core should be derived from your internal description, not typed alongside it.

In practice

  1. Use the fifteen elements as a completeness checklist.
  2. Describe in a rich model, and project to Dublin Core on export.
  3. Prioritise identifier, rights, and creator: the most commonly missing and the most expensive to reconstruct.
  4. For an institutional deposit, plan for LIDO or CIDOC CRM as well.

Obscura Flow describes against a full heritage model — work, print, attribution, provenance — and exports to LIDO 1.1 and CIDOC CRM; Dublin Core-style elements are derived from it rather than retyped.