Duplicate Image Replacement in Melbourne's Digital Archives: The Key Decisions Ahead

Melbourne's major public institutions are facing a crunch point over how they manage tens of thousands of duplicate images stored across ageing digital asset systems — and the decisions made in the next six to twelve months will shape how Victorians access visual records for years to come.

The issue has been building quietly inside heritage bodies and local councils for several years, but it has sharpened recently as the Victorian Government pushes institutions to consolidate infrastructure ahead of a broader Digital Public Records Strategy review scheduled for late 2026. Duplicate images don't just waste storage — they create legal exposure around licensing, confuse provenance records, and make it harder for the public to trust what they're searching for online.

Where the Problem Is Most Acute

The State Library Victoria on Swanston Street is among the institutions managing the largest public image repositories in the state. Its digitised collection spans millions of items, and library administrators have acknowledged publicly in past annual reports that deduplication across cataloguing systems remains an ongoing technical challenge. The library's current digital preservation framework, last updated in 2023, references the complexity of merging records from legacy databases with newer platforms.

The City of Melbourne, whose digital asset library covers everything from planning photographs taken along Flinders Lane to event imagery from Federation Square, faces a parallel problem. Council staff responsible for records management have to navigate both copyright clearance questions — particularly for third-party photographs acquired over decades — and the practical question of which version of a duplicated image carries the authoritative metadata. Get that wrong, and you risk publishing an image with the wrong attribution or an expired licence.

The Australian Centre for the Moving Image in Federation Square has been further ahead on this than most, having completed a partial deduplication audit of its born-digital collection in 2024 as part of a broader collection management overhaul. Even so, the process required dedicated staff time and purpose-built software to flag near-identical images — files that are visually the same but saved in different formats, resolutions or file names.

What Happens Next — and Who Decides

The core decisions ahead fall into three categories: technical standards, governance, and cost allocation.

On the technical side, institutions need to choose between perceptual hashing tools — software that identifies visually identical images regardless of file format — and more labour-intensive manual audits. Perceptual hashing is faster and cheaper at scale, but it can flag culturally significant variants as duplicates when in fact subtle differences in a cropped or colour-corrected version carry archival meaning. For a body like the Public Record Office Victoria, based in North Melbourne, that distinction matters enormously.

Governance is thornier. Who gets to decide which version of a duplicate image becomes the canonical record? In a siloed institution, that's a curator's call. Across shared platforms — and the Victorian Government has been pushing for more shared digital infrastructure since at least the 2022 Digital Strategy for the Public Sector — it requires cross-agency agreement that doesn't yet exist in any formal policy document.

Cost is the most immediate pressure. Commercial digital asset management platforms that include built-in deduplication tools typically run anywhere from $30,000 to $200,000 annually depending on collection size, according to publicly available vendor pricing from suppliers active in the Australian market. For smaller councils and regional galleries already managing tight operational budgets, that's not a trivial line item.

The Victorian Government's Department of Government Services is expected to release updated guidance on digital records management standards before the end of 2026. Institutions that wait for that guidance before committing to a deduplication approach may find themselves behind the curve — but those that move too early risk locking in systems that don't align with whatever framework emerges. The smarter play, according to the experience of institutions that have already run pilots, is to begin internal audits now, document what you find, and treat any vendor procurement as modular enough to adapt. The images won't sort themselves.