What makes a listed building different to map?

It cannot be altered, marked or rigged to. On a set you can add registration targets, move a flat or repaint a surface; on a listed building none of that is available. Everything has to be achieved from the projector side, which puts all the difficulty into calibration and content.

How are compound curves mapped?

With a dense mesh derived from a survey or scan rather than by dragging vertices by eye. The shells of a building like this curve in two directions at once, and there is no flat region anywhere to anchor to, so architectural work almost always begins from measured geometry.

Where can the projectors go?

Wherever the site permits, which is usually not where the design wants. Distance, angle and available power are all fixed by the surroundings, and long throws mean large, bright, heavy machines with correspondingly serious rigging and weather protection.

Does weather affect architectural projection?

Substantially. Outdoor projection surface work has to survive wind loading on the projector positions, moisture, and ambient light from the city itself, which raises the black level and reduces effective contrast far more than most indoor work ever has to deal with.

Is content different for a building?

Yes. A building has meaning and a form the audience already knows, so content that ignores the architecture reads as wallpaper. Work that succeeds tends to respond to the shells (following their edges, using their divisions) rather than treating them as an inconveniently shaped screen.

Sydney Opera House as a Surface

A single large curved architectural surface with a fine grid of magenta light warping across it. — A mesh derived from survey, not from dragging corners

Projecting onto the Sydney Opera House belongs to a different discipline than touring work, even though the technique is nominally the same. A stage set is designed to be projected onto. A listed building is designed to be a building, and it is not going to accommodate you.

Geometry you have to measure, not invent

On a set, the model exists before the object; the build follows the drawing. Here the object came first by half a century, and the model has to be recovered from it by survey or scan.

The shells curve in two directions simultaneously and there is no flat region to anchor a warp to. Vertex-dragging by eye (the standard on-site method for scenic work) is not viable at this scale or on this geometry, so architectural jobs are calculated first and verified second.

Projector positions the site chooses

Throw distance, angle and power availability are all decided by the surroundings. Long throws demand very bright, very large machines; those machines need structure, weather protection and a power supply, all of which have to exist somewhere the public is not.

The resulting positions are frequently steep or oblique, which pushes more correction into the mesh and costs effective resolution on the surfaces furthest from the lens.

Large smooth curved white shells overlapping one another, a hard light raking across them from one side so each curve turns from bright to black along its length. — Compound curves · geometry that has to be measured, not invented

A building of this kind cannot be modelled from a drawing. The surfaces are compound curves that overlap and self-shadow, and a warp calculated against an idealised version of them lands visibly wrong along every edge where the real shells differ from the drawing.

So the geometry has to be surveyed. That is slow, it happens before anything creative starts, and it is the single largest difference between mapping a building and mapping a set somebody built to a specification.

Content that acknowledges the building

The distinctive failure of architectural projection is content that would have worked equally well on a wall. A famous building carries meaning and a form the audience already holds in mind, and imagery that ignores it reads as decoration applied to something more interesting than itself.

What tends to work is content that treats the structure as a participant: following the edges of the shells, using their divisions as compositional boundaries, letting the form lead. The same instinct governs mapped scenic work generally (see projection mapping) but here there is no option to change the surface if the content does not fit it.

Where the technique went next

The scale is unusual; the problem is not. It is the same discipline that the mapped cube exercises at a few metres (surfaces that turn away from the lens, seams that only resolve from one position) with the added constraint that nothing about the object can be changed to make it easier.

The technique is set out under projection mapping; where the projectors can physically stand is a structural and positional question, resolved with the throw-distance calculation and modelled in previs before anyone travels.

How this page is sourced

What is recorded: the design intent, the arrangement, the technical problem it set out to answer, and the consequences for touring. What is not recorded: fixture counts, equipment models, lumen figures or content specifications, which appear in trade coverage inconsistently and are not reproduced where a single source cannot be identified. Where the published record is thin about survey method, this page says so rather than filling the gap.

Published coverage of the work is gathered under the field in print, listed with publisher and date. Registration and previs carry narrower meanings here than in general use and are defined in the glossary. The editorial basis for the page, including what is done when the record contradicts itself, is set out on about this site.

Sydney Opera House as a projection surface: mapping a listed building