The (Roman) Art of (Computer) Science: 3D Computer Technology and a Recreation of the "House of the Drinking Contest" at Antioch

John J. Dobbins, professor of Roman art and archaeology, University of Virginia, and Ethan Gruber, Web Applications Developer, University of Virginia Library/3D Modeler, Pompeii Forum Project

Among the treasures discovered by the 1930s excavations at the ancient Roman city Antioch-on-the-Orontes (in modern-day Turkey) and its vicinity were numerous polychrome floor mosaics. One of these mosaics, depicting the drinking contest between Dionysos and Herakles, is one of the masterpieces in the Princeton University Art Museum's collection of ancient art. The mosaic originally decorated the pavement of the dining room in the eponymously named "House of the Drinking Contest," a third-century house in Seleucia Pieria, the port of the Roman city of Antioch-on-the-Orontes. The drinking contest and five other associated pavements were distributed to six museums, all but one in the United States.

The placement of the mosaics in museums has severed them from their original site and from each other. Excavation photographs show the mosaics in situ, but the absence of the original roofs and supporting walls altered the lighting conditions substantially. The ancient occupants would have seen sunlight as it entered the house over walls, through windows, and between columns, and moved across pavements during the course of the day and the arc of the year.

Computer science addresses the problem of the lost architectural and lighting contexts through a three-dimensional model that recreates the house and inserts the mosaics into the spaces that originally contained them. This computer model thus presents a graphic rendering of the hypotheses underpinning the architectural reconstruction. A programming script, which calibrates the sunlight to the latitude and longitude of Seleucia Pieria in the year 230 A.D., enables an accurate lighting simulation for the house and its mosaics. This simulation presents the lighting conditions at the summer and winter solstices of that year, and time-lapse videos allow the viewer to observe and study the movement of light throughout the house. Finally, the model places the viewer within the ancient spaces, thereby reconstructing and recontextualizing views from corridors within the house and from the house to the natural environment