Spent the last fortnight on a sandstone restoration job on a late-Victorian building in Sydney’s inner suburbs — replacing two badly weathered cornice sections and two damaged corbels. It’s the kind of work I’ve been doing for 25 years. The way we now do it is meaningfully different from how we did it 15 years ago. Worth writing down what’s changed.

The biggest shift is documentation. Photogrammetric scanning of damaged ornament has become standard practice on any serious heritage job. Before any stone is removed, the existing ornament is captured in three-dimensional detail at high enough resolution that the geometry is fully recoverable. The scan data drives the replacement carving, the conservation report, and the final QC against the original profile.

This sounds technical and it is, but the practical implication is simple: we’re now able to replicate weathered original profiles with a degree of accuracy that wasn’t possible when measurement relied on rubbings, manual gauges, and photographic reference. The work that gets installed today is more faithful to the original than what was being installed in the same buildings during 1990s and 2000s restoration cycles.

The CNC question

Whether to use CNC roughing for replacement ornament is now a settled question on most jobs. Yes, where the geometry is regular enough that a five-axis machine can produce a reliable rough form, and the carver finishes the work by hand from there. The quality benefit is that the carver spends their time on the work that requires judgement — the surface texture, the historic tooling marks where authenticity demands them, the final character of the piece — rather than on the labour of removing bulk material to a known profile.

The cost benefit is real but isn’t where most people think. CNC roughing on a complex ornament might save 30 to 50 per cent on the rough-out time. It rarely changes the finishing time, which on a serious piece is most of the total time. The economics make sense for production-volume work — restoring a long balustrade with repeating elements, replacing a series of cornice sections — but for one-off bespoke work, the setup overhead can outweigh the savings.

Where I draw the line is on small ornament where the historic character is dominated by hand tooling — capitals, foliate work, figurative details. Roughing these by CNC produces a result that the trained eye can see, and that the building heritage authorities can sometimes refuse to accept. Hand-cutting these from rough block is slower and more expensive, but it’s the right answer for the right buildings.

Stone matching has gotten harder

The sandstone supply situation in Sydney has steadily worsened over the past 15 years as quarries closed, were heritage-listed, or moved to non-architectural production. Matching original sandstone for a Sydney heritage restoration in 2026 is harder than it was in 2010, and the gap is widening.

The practical responses are imperfect. Some jobs use sandstone from currently-producing Australian sources (Pyrmont where it’s still available, Helidon, certain Queensland quarries) and accept that the colour and grain match isn’t perfect. Some jobs source recycled stone from demolished buildings of the same era. Some jobs use limestone alternatives and document the substitution in the heritage report.

None of these are great. The best answer for a serious heritage job remains stone reclaimed from buildings of the same era, and the supply of that is finite and getting rarer. Heritage authorities are beginning to ask harder questions about substitute materials, and the conversations during permitting are now where the major specification decisions are being made.

For more complex restoration projects with structural integration requirements — incorporating modern engineering reinforcement into heritage stonework, integrating contemporary services into historic fabric — many of the larger jobs now bring in specialist consultancy for the structural and project management piece. The carving work is what we do; the project integration around it has become its own discipline.

Conservation versus restoration

The conservation versus restoration conversation has shifted toward conservation more strongly than it had 10 years ago. Where original stone is sound enough to retain, the current preference is to retain it even with some weathering damage, rather than replace it with new ornament that’s geometrically faithful but materially different.

This is the right answer in principle. In practice it produces different outcomes that clients sometimes find unsatisfying. A conserved cornice with selectively replaced sections looks visibly older than a fully replaced cornice. The historic authenticity is preserved; the visual unity isn’t.

The conversations during commissioning need to address this directly. Clients who want their heritage building to look brand-new are usually asking for restoration that crosses into reconstruction. Clients who want their heritage building to look like a heritage building need conservation, which means visible age, which means the replacement work should sit alongside the original rather than replacing the original wholesale.

Workmanship documentation

The other significant change is in workmanship documentation. Photographs of every stage of the work, daily work logs, dimensional records, materials provenance certificates — what would have been a single page of paperwork on a 2010 job is now a detailed dossier on a 2026 job. The administrative load is real and is reflected in the project costs.

The benefit is that the work is genuinely defensible against later questions. Heritage councils and insurance providers ask harder questions than they did, and the documentation that supports good answers takes less effort than the documentation that supports defensive answers.

What I tell apprentices

The trade is harder to enter than it used to be and the work is more demanding. The hand skills are still the foundation. The digital skills — photogrammetry, CNC operation and finishing, dimensional documentation — are no longer optional. The communication skills with heritage authorities, project managers and clients are increasingly part of the job.

The apprentices who’ll have the strongest careers in 2030 are the ones who treat the trade as a craft that includes contemporary tools, not a craft that resists them or a contemporary process that pays lip service to the craft. Both extremes produce inferior work. The integration is where the good restoration is happening.