Visualising Roman Canterbury: Computer Graphics in Archaeology


Canterbury temple precinct from the air, looking west

Computer-generated graphics have become commonplace in many areas of modern life. Few can have failed to notice their impact on the work of the television, film, advertising and publishing industries. Elsewhere, engineers, architects and others have turned to Computer Aided Design (CAD) systems in preference to the traditional drawing board. Perhaps less widely appreciated is the use of many of these techniques in visualising and exploring past environments.

Canterbury temple precinct, looking west

Computer graphics have long been used in archaeology and in museums for everyday tasks such as recording excavation plans, illustrating artefacts and presenting the results of scientific analyses. Today, they are also increasingly used both to help archaeologists examine possible 'reconstructions' and in museums as a means of presenting this information to the public.

Canterbury temple precinct, looking north-west

The illustrations accompanying this article come from projects aimed at examining the use of computer graphics in archaeology and, in particular, attempting to visualise how parts of roman and medieval Canterbury may have appeared to its inhabitants. Some of this work has been carried out in collaboration with the Canterbury Archaeological Trust and a simple display was produced for Canterbury Museums in 1992. This enables visitors to the new Roman Museum in Butchery Lane to see a 'reconstruction' of one of the buildings uncovered in the Trust's 1990 excavation of the Longmarket.

General view of the Longmarket bath house

Producing these pictures is a two stage process: modelling and rendering. The modelling stage creates a geometric specification of the objects to be viewed, and the rendering stage converts the model into a picture of the objects seen from a chosen viewpoint. The model can be in the form of lines representing the outlines of simple shapes, polygonal shapes (usually triangles) defining their surfaces, or combinations of simple solid shapes such as cubes, spheres, cylinders and cones. The latter 'solid modelling' approach has been used here because these simple solids are ideal for representing the forms of classical architectural.

Sectional view of the Longmarket bath house (108kB)

Once the preserve of well-equipped research laboratories, impressive results can now be achieved on desktop PCs. Both free and commercial software is available for both modelling and rendering. The images shown here were produced using the freely available POV-Ray program which uses the 'ray-tracing' method of rendering. Ray-tracing of complex models may take many hours to complete, but it is widely used because it provides more realistic lighting than many of the faster methods.

Roman temple (after Vitruvius) (76kB)

Models may be generated using a CAD or other modelling program as a means of drawing and placing the basic shapes of the buildings, although the model of a roman temple illustrated here was generated automatically by a computer program. This program uses a set of rules, derived from the writings of the 1st century AD architect Vitruvius, that define the relative proportions of the many parts of a temple. The only input required from the user is the number of columns across the front and the height of a column.

North-west corner of the temple precinct (51kB)

The building models are based on the available evidence from excavations and the interpretations placed on that evidence by archaeologists and architectural historians. Usually this means starting from a plan of incompletely surviving foundations although, in some cases, the remains may be recognizable as a well-known form of building such as a temple. More often, though, the remains are fragmentary and may not give a clear impression of the extent or form of the building. Here, the interpretation is necessarily speculative, guided by specialised knowledge of the range of possibilities for the architecture of the period and the materials available in the region.

Shrine in temple precinct (53kB)

The technology certainly produces visually impressive results, but are they anything more than pretty pictures? Do they help the archaeologist to arrive at better interpretations? Do they offer the museum visitor a more rewarding experience? The main limitation of drawings and paintings is that they present a single fixed view. It is all too easy for an archaeologist to miss problems in their reconstruction such as walls that do not join, or an unsupported roof. These issues must be confronted and solved when building either physical scale models or computer models.

Inside the tepidarium of the Longmarket bath house (114kB)

In a museum context, scale models can allow the visitor to choose their own viewing angle, but they are often limited by their scale. Except in rare full-size examples, only a 'bird's eye' view is possible and one cannot walk around inside the model and so gain any impression of how the building may have appeared to its contemporary users. Scale is not a problem in a computer model. Any viewpoint is possible and the visitor can be presented with static views, video tours and, increasingly, interactive walk-throughs where the user chooses their own route through the model.



Last updated: 20th January 1997


100% Dublin Core metadata (select View | Document Source to see the metadata)