Duplicate Images in Melbourne's Public Art Register: What Happens Next and the Key Decisions Ahead

Melbourne's public art and cultural heritage records contain hundreds of duplicate images — photographs catalogued under multiple entries, works misidentified across separate databases, and digital files orphaned from their original metadata. The problem, known inside archival and local government circles as duplicate image replacement, has moved from a bureaucratic nuisance to a live policy question as the State Government presses ahead with a broader digitisation push across Victorian cultural institutions.

The timing matters because money is now on the table. The Andrews-era Victorian Cultural Infrastructure Fund has been succeeded by a successor grants program administered through Creative Victoria, with several councils and collecting institutions currently mid-application for digitisation support. How duplicate image problems get resolved — and who bears the cost — will shape the integrity of those new digital collections for a generation.

Where the Problem Lives in Melbourne

Two institutions sit at the centre of the immediate dispute. The City of Melbourne's Public Art database, maintained through the council's Arts and Culture branch and accessible via the Melbourne Art Map, has been flagged internally for containing overlapping image records tied to the same physical works. Separately, the State Library of Victoria on Swanston Street, which manages one of the country's largest photographic holdings, is partway through a multi-year project to reconcile its digitised collections — a process that has exposed recurring duplicate entries dating back to early scanning campaigns in the late 1990s.

In the inner north, the Darebin City Council has its own local heritage photographic archive, housed partly through a partnership with the Northcote Library on High Street. Council staff there have been working with a third-party digital asset management vendor since early 2025 to decommission redundant image files and rebuild metadata tagging from scratch. The project, originally scoped to wrap by June 2026, is running behind schedule, according to council documents tabled at the April 2026 ordinary meeting.

The core technical question — replace the duplicate with a canonical master file, or merge records and preserve all variants — sounds simple. It is not. Archivists argue that even a poor-quality duplicate can carry provenance value: a scan made in 1998 might show a work before it was restored or damaged. Deleting it means losing that forensic layer permanently.

The Decisions That Will Define the Outcome

Three choices now sit in front of administrators, and the answers will determine how this plays out across Victoria's collecting network.

First, the metadata standard question. Creative Victoria has been consulting on adopting the Dublin Core Metadata Initiative framework across state-funded collections since mid-2025. If that standard is mandated — which a decision expected by September 2026 would confirm — institutions will need to retrofit existing records, including resolving duplicate image conflicts before migrating. The cost of that work is not trivial: comparable projects in New South Wales have run to tens of thousands of dollars per institution for collections of moderate size.

Second, the deaccessioning authority question. Currently, no single body has clear authority to instruct a council to delete a duplicate image from a jointly managed database. The Victorian Public Record Office sets retention standards under the Public Records Act 1973, but digital image files occupy an ambiguous space in that legislation. A review of the Act's application to born-digital and digitised cultural assets was flagged in the 2024-25 state budget papers but has not yet produced a public consultation draft.

Third, the resourcing question. Most of Melbourne's 31 councils do not employ a dedicated digital archivist. The Australian Society of Archivists has identified that gap as a structural vulnerability across local government collections nationally. Without in-house expertise, councils outsource, and outsourced projects — as Darebin's experience illustrates — frequently slip on timeline and scope.

The most pragmatic path forward, based on how similar programs have resolved overseas, involves a staged triage: flag exact duplicates for suppression rather than deletion, preserve near-duplicates with variant metadata, and establish a central audit log so future researchers know what was removed and why. Whether Melbourne's institutions have the budget, the staff, or the political will to follow that path before the next digitisation funding round closes in late 2026 is the question no one has yet answered.