Digital casts and 3D models: new ways of looking at Roman Emperors on coins

Lire ce billet en français : Moulages numériques et modèles 3D : de nouvelles façons de visualiser les empereurs romains sur les monnaies

The peculiar nature of coins as small, portable and serially produced metal objects has always presented unique challenges to scholars and museum curators, who face the task of cataloguing and studying sheer numbers of collection entries as well as of old and new finds, while striving to make them accessible to as many people as possible. Coins are also among the most underappreciated classes of archaeological materials in museum galleries due to the difficulties of combining adequate lighting and display solutions, so another major task is to give visibility to very small, either too dark or too shiny objects, and to make them appealing to the general public.

Fig. 1 3D scanning with a blue-laser scanner of the bronze head of Augustus from Meroe at the British Museum for the RESP project

A huge help to tackle all these challenges has come, especially in the last fifteen years, from digital technology, which has found many applications to the study of the ancient world, including coinage, allowing scholars to speed up and automatise all the most common activities in the numismatic practise.

The greatest progress has been made in the digitisation process for identifying, cataloguing, photographing, archiving and publishing ancient coins, but much more advanced technologies, like Computer Vision and Artificial Intelligence, have also been deployed in complex operations such as the study of die-links and sequences, which used to require exhausting and endless manual work and can now be performed in much shorter time on a PC. These resources have also changed how coins can be visualised and appreciated in every detail by the general public on screens and using interactive apps, both in museum displays and from home or in special learning programs at the university.

One branch of technology that has contributed substantially to this digital revolution is three-dimensional imaging, which has become very popular in Archaeology for the many opportunities that it offers to study, analyse and virtually reconstruct ancient artefacts, environments and even entire monuments (fig. 1). Applied to Numismatics, it offers the possibility to acquire images of coins in 3D, incredibly high-resolution point clouds reproducing their shape, volume and geometry, emphasising every feature of their structure with micrometric precision (i.e. under a millimeter). 3D models of coins can be enlarged, flipped, rotated and analysed in greater detail than from digital photographs, so they can be used in research as much as in dissemination and outreach activities (fig. 2).

Fig. 2 3D scanning with a microprofilometer of a Roman provincial coin from the Castelvecchio Museum (Verona) for the RESP project

The University of Verona (Department of Cultures and Civilisations), in collaboration with King’s College in London and the WMG at Warwick University, is conducting a large campaign of 3D scanning of Roman imperial and Roman Provincial coins as part of RESP (The Roman Emperor Seen from the Provinces), a five-years research project funded by the European Research Council under the Horizon 2020 Program (GA 101002763). RESP studies the genesis, replication, and diversification or portraits of Roman emperors and their family members in the provinces during the first three centuries of the Empire (c. 31 BC – AD 297). It aims to get a better understanding of how portraits were designed and produced, how they differed from the ones made in Rome and why.

The project is combining traditional research methodology with three-dimensional imaging especially to reconstruct how the portraits of the same emperor might have originated from a shared model and were reproduced and adapted on different media, with a major focus on the comparison between coins and sculptures. Coins, in particular, were the main medium for spreading the imperial image and they are our prime source for observing how imperial portraits changed over time and from the issues produced in Rome (and other state mints) to the ones produced by hundreds of provincial mints; they carryed miniaturized versions of the official portraits designed by the central authority for sculpture, which circulated in every corner of the Roman world and were copied by local artists and engravers, generating provincial adaptations.

High-definition scanning of coins enhances the legibility of facial features, allows for magnification and reduces the degree of potential subjectivity in the identification of portrait typologies, as the captured point cloud can be used as an objective, measurable frame of reference for comparison between official models and their copies. In the project we 3D scan portraits in sculpture and coinage to see them together in three-dimensions: the portraits of the same emperor in different media and formats can be scaled together, aligned, overlapped and compared to understand their possible dependencies from a shared model as well as their discrepancies to reconstruct their genesis and production.

Fig. 3. The process of virtually generating a portrait of Hadrian in the round from a 3D scan of a coin profile with a Morphable Model

We have also developed a digital application designed specifically for our research, which uses 3D modelling based on a Morphable model for 3D human faces to generate portraits in the round from the coin profiles. This form of digital experimental archaeology reenacts how a prototype could be modified and adapted to fit into different formats, providing insights into the maker’s perspective and highlighting the technical differences between the processes of engraving and that of carving (fig. 3).

Besides this very focussed application, 3D imaging is used in several other branches of numismatic research, for instance to study how coins were made and distinguish authentic specimens produced by striking from copies or fakes made by casting, or to analyze specific characteristics in detail, for instance intentional modifications of coin designs and legends, such as countermarks and graffiti.

Three-dimensional images can be created with many different techniques. The most common one is photogrammetry, which consists in taking multiple pictures of an object from 360° in order to stitch them together and create a digital model using SfM (Structure from Motion) algorithms. This is the most accessible method, because it is easier to perform and it relies on cheaper equipment (consumer grade digital cameras), but it has some practical limitations, entailing that either photographs are taken by moving the camera around the coin to capture its entire surface, or that the coin is rotated by a certain angle before every shot; furthermore, the results are not always satisfactory. 3D scanners using structured light or blue laser technology can produce better results because they are more efficient and faster, but require larger and bulkier equipment. Conversely, mainstream portable and handier commercial scanners, regularly used in Archaeology, are mostly unsuitable for coins, because they are designed for much bigger objects and do not have enough resolution to capture the finer details. The accuracy of the scans also depends, as in digital photography, on the peculiar characteristics of coins’ metal surfaces, which can be either too reflective or too light-absorbing: most portable 3D scanners can acquire images of gold and silver coins, but they struggle with specimens made of copper alloys, especially darker patinas, which form a huge class of materials, extremely common from the Hellenistic period and especially in the Roman world.

Fig. 4 The 3D scanning of a bronze Roman provincial coin of Nerva from Cassandrea (Macedonia) at the BnF (FG 601) with Elastomeric Tactile Sensor technology

Conversely, to scan a sample of Roman imperial and Roman provincial coins at the BnF as part of the RESP project we used a different instrument called GelSight Mobile, which works in a different way. It is a small portable device that uses elastomeric tactile sensor technology directly on the coin surface to take an impression of the relief: the soft surface of the sensor adapts to the shape of the coin and captures multiple shots from different angles applying a technique called Photometric Stereo reconstruction (fig. 4). The software then uses these images to instantly reconstruct the 3D geometry of the coin with micrometric resolution (fig. 5). Besides being extremely fast and precise, the device works efficiently on any surface material, regardless of composition, reflectivity, transparency, or ambient lighting conditions, without affecting the object’s integrity of leaving traces on it.

Fig. 5 Digital picture of the coin of Nerva (left) next to the 3D scan of it (right)

In the old days numismatists used to make casts of coins in order to have an exact reproduction of their die impression. Made of plaster or modelling clay, they served better than photographs in rendering the details of coin designs and legends; hundreds of them were patiently collated to assemble corpora of specimens from different collections to reconstruct die identities and die sequences in studies focused on an ancient mint or issuing authority, like a king or an emperor. The way the tactile sensor works recalls the same principle, producing a digital cast of the coin which can be repeated multiple times without affecting the object and is stored automatically in a digital format that eliminates risks of deterioration. Scans can be tailored for specific needs, like extracting data of one particular portion of a coin and generate a perfect copy of it in three dimensions: for this reason it suits the purposes of the RESP project particularly well as a mean to acquire high-resolution images of imperial portraits. The basic function of the device is to provide images that can be enlarged, rotated and observed from different angles to analyze details such as curls, beard, eyes and all the key features that conform to a certain portrait typology. Even though the scan lacks color information, the 3D surface can be viewed in plain monochromatic versions that allow for an objective inspection of its peculiarities (fig. 6).

Fig. 6 3D scan (left) and 3D surface (right) of a gold aureus of Nerva (BnF IMP-9405 / BNC III Nerva 45)

Among the coins that we scanned in the BnF collection, a special mention goes to the exceptional gold aureus struck by the mint of Cologne in AD 267 for emperor Postumus (AD 260-269), one of the 3rd century usurpers who established an independent enclave known as the Gallic Empire. The coin displays one of the earliest examples of an incredibly rare typology of portrait, which is seen from a three-quarter view instead of from profile (FG 1399). This specimen stands out also for the extraordinary quality of the engraving that emphasizes every single detail of the emperor’s lineaments, which are documented exclusively by his coinage. The 3D scan brings out the exceptionally accurate rendering of the hair and beard, as well as the differences in height of the relief between facial features that are normally much more even in the profile view, such as the eyes and especially the nose, which reaches the highest peak of protrusion (fig. 7).

Fig. 7 3D surface of a gold aureus of Postumus: obverse (FG 1399)

The software allows to appreciate all these details by using a horizontal slider that controls the vertical (Z-axis) scale of the model, exaggerating depth differences to make the relief more perceptible and easier to analyse: on the reverse, which depicts a mythological scene, Hercules taming one of Diomedes’ mares, we see the different treatment of the design between the animal’s body in very low-relief and that of the hero in higher relief in the foreground, to emphasize the depth of field of the entire scene (fig. 8).

Fig. 8 3D scan (left) and 3D surface (right) of a gold aureus of Postumus: reverse (FG 1399)

Additionally, the device’s software, designed to be used for metrology and quality control in large industries (Aerospace, Automotive, Robotics), allows to analyze the morphology of the coin and generate visual and written reports of the results. They include graphs of the profile geometry of the die impression and the offset between its different parts (fig. 9); each distance can be measured and recorded to make comparisons between similar designs or the same design struck by different dies. In fact, other than low-relief surfaces, this technique can be used to analyze also engraved materials that present shallow depressions on the surfaces, such as small stone gems, especially intaglios, and metal dies used to mint coins in antiquity, where, besides the design, micrometric damages or fractures, or simply signs of wear can be detected, measured and analyzed.

Fig. 9 Graph showing the offset of the 3D scan of a coin-die of Augustus from Nîmes (FG 2396)

See for instance the well-known conical bronze die used to strike the obverses of some emperor Augustus’ aurei, which was found in 1739 inside a fountain in Nîmes (FG 2396; fig. 10).

Fig. 10 Bronze coin-die of Augustus from Nîmes (FG 2396)

The scan allows to visualize the die from two points of view: not only the concave surface of the engraved side but also its opposite, the convex one corresponding to the relief that was stamped on coins in the minting process, an image that can be compared and overlapped with that of actual specimens originating from it (fig. 11).

Fig. 11 3D surface of the die showing the engraved profile (left) and the relief profile (right)

The software also generates heatmaps of the scanned surface, marking with different colors the varying height of the die surface, degrading from colder tones of the lowest points to warmer of the highest ones (fig. 12).

Fig. 12 Heatmap of the 3D surface of the die

If this technology becomes more accessible and widespread, it certainly has the potential to have a great impact not only on how researchers study and present coins to the academic community, but also on how the wider public can appreciate their importance and learn from them.

Dario Calomino

To find out more:

OpenEdition vous propose de citer ce billet de la manière suivante :
dcalomino (17 juin 2025). Digital casts and 3D models: new ways of looking at Roman Emperors on coins. L’Antiquité à la BnF. Consulté le 21 janvier 2026 à l’adresse https://doi.org/10.58079/14560


Laisser un commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *

This site uses Akismet to reduce spam. Learn how your comment data is processed.