The app recovers the lost church of San Pier Maggiore as a digital environment that can be experienced via Augmented Reality (AR) both in its location in Florence and in the Sainsbury Wing of the National Gallery in London.
Building on five years of research, the current app presents a digital visualization of one of Florence’s most important churches, formerly home to an array of artistic masterpieces but demolished at the end of the eighteenth century. In 2015, a team from the University of Cambridge and the National Gallery, London conducted fieldwork to collect new evidence for San Pier Maggiore’s physical fabric preserved inside later buildings on its site. Previously only the church’s seventeenth-century façade was thought to survive. By stitching thousands of photographs together using Autodesk ReCap 360 they created a virtual neighbourhood and were able to better define the layout of the church interior using historical plans and sections, adding structural elements that are now embedded in today’s architecture. The resulting film, directed by Miguel Santa Clara, includes a point-cloud visualisation of the medieval church nested within the contemporary cityscape and can be viewed here.
In 2017-18, researchers from the Universities of Exeter, Cambridge and Florence, supported by the Getty Foundation, returned to San Pier Maggiore to refine the model of the church. The aim was to exploit the potential of AR to communicate to a wide audience the layout of the church interior and the interrelationships between its architecture and artworks. The resulting app – realized in collaboration with app developers at Calvium and AR specialists at Zubr.co – allows users to walk through the space, exploring the 40-metre long nave and engaging with its architectural features. The model can be anchored either in Florence on the site where the church once stood, or in the Sainsbury Wing of the National Gallery in London, orientated around Jacopo di Cione’s majestic 1371 high altarpiece for the church depicting the Coronation of the Virgin. Standing before the painted panels in the gallery, image recognition technology enables users to access the model and to visualise the altarpiece’s original location within the church. In Florence, a geo-fenced area locks the altarpiece to the site where it once stood, making the model available for on-site navigation. The street layout lines up with the model, and structural features of present-day shops and residential buildings can be understood as elements of the lost church, embedded into today’s streetscape.
The model that you experience in the current version of the app integrates different grades of evidence and certainty.
It incorporates elements for which we have direct surviving evidence, in the form of physical traces of the original building, historic plans and representations. We can confidently reconstruct the ground-plan and dimensions of the church, its principal building phases, its basic materials (like the reddish-grey pietra forte stone), and the locations of its altars and chapels. The modelling of Jacopo di Cione’s high altarpiece derives from detailed research undertaken by the National Gallery and art historians on the surviving panels. This evidence can be supplemented by extensive documentary and archival evidence. While the physical fabric of San Pier Maggiore does not survive, its archive does, transferred in its entirety to Florence’s state archives. Ironically, this vanished church is one of the best documented in the city.

The app model also draws on evidence from comparable buildings, notably other Florentine churches built in the same period. Like San Pier Maggiore, the much larger church of Santa Croce, begun in 1295, was under construction in the early years of the fourteenth century and shares many of its architectural features. The model draws on Santa Croce, for example, for the rendering of the interior doors on the façade. We know that San Pier Maggiore possessed an open beam roof and here the model is informed by surviving wooden roofs in other Tuscan churches of similar size and date. The terracotta-tiled floor in the model is typical of pavements from the late medieval and Renaissance period in other ecclesiastical buildings in Florence.
There are also details that are documented but which are missing from the current model, some of which we hope to add to future versions. There are, for example, other surviving altarpieces from the church dispersed amongst different museum and gallery collections. For now we have included bare altar blocks where we know that there were side altars to indicate that the church was not an empty space; in the future these altar tables could be populated with their documented artworks. Similarly, we have detailed information for tombs and coats of arms in the church interior which could be added to the model. Conversely, we have almost no evidence for its fresco decoration, which may have been extensive, and this will be much harder to visualize. The choir stalls and dividing screen that we know were in the central nave are currently missing and can only be visually reconstructed in outline terms.

These issues of uncertainty and authenticity are inherent in any digital visualization of historic environments. For a building like San Pier Maggiore there is an added challenge of selecting a particular moment in time to visualize when the church interior was in a constant state of change and renewal. This aspect of the building’s history would need a time-slide feature to adequately articulate. The present model, however, does include accretions over time to the original early fourteenth-century building, like the Renaissance side chapels with grey pietra serena columns and arches which were added to the church in the late 1400s.
While the models strives for accuracy, our knowledge of the historic fabric will always be imperfect, and its digital visualization should never be understood as an unproblematic recreation of the past. Rather it is closer to a dialogue between past and present, but one with the potential to reshape our experience of both historic artworks and contemporary cityscapes.
Donal Cooper and Chiara Capulli