Navigation – Plan du site

AccueilNuméros51III. Mise en œuvre, polychromie e...Renaissance plaster casts: first ...

III. Mise en œuvre, polychromie et conservation

Renaissance plaster casts: first insights by means of Structured Light 3D Scanner

Les moulages en plâtre de la Renaissance : premières données fournies par le scanner 3D à lumière structurée
Gianluca Gariani, Charlotte Hochart et Nicolas Mélard
p. 105-117

Résumés

Dans le cadre d’un programme d’étude scientifique des reliefs de dévotion florentins du xve siècle, leur production en série par le biais de moulages a fait l’objet d’une recherche ciblée. Douze œuvres ont été examinées afin de mieux comprendre les techniques de fabrication des stucs appelés Madonne di gesso. Il s’agissait de moulages d’après trois modèles, créés respectivement par Donatello assisté de Bartolomeo Bellano, Desiderio da Settignano et Antonio Rossellino conservés au Louvre et dans deux musées florentins (Bargello et Bardini). Grâce à plusieurs campagnes d’études par scanner 3D à lumière structurée des informations précieuses ont été livrées aux historiens de l’art. L’examen des modèles et la comparaison de maillage de leurs modèles 3D ont permis de distinguer et de mesurer des différences millimétriques dans les formes et la topographie de reliefs d’une même série de moulages ou d’après un même modèle. Il existe d’ailleurs des corrélations directes établies par l’examen de bas-reliefs vraisemblablement issus du même moule ou prototype. Dans certains cas, nous proposons une chronologie au sein de la série de moulages. Les résultats donnent des éléments de compréhension pour la fabrication de ces œuvres dans les ateliers florentins du xve siècle.

Haut de page

Texte intégral

Introduction: the serial manufacturing of Renaissance devotional reliefs

  • 1 Gentilini, 2012; Cat. Exhib. Florence-Paris, 2013, p. 426-429, VIII.2-3.
  • 2 Vasari, 1550 (ed. 2008); Gaurico, 1504 (ed. 1969); Biringuccio, 1540 (ed. 1990); Lebrun et al., 188 (...)

1In 15th century Florence, renowned Old Masters’s workshops (e.g. Donatello, D. da Settignano, A. Rossellino) produced outstanding examples of devotional reliefs portraying the Madonna with Child in different iconographic variations (i.e. types or models1). The peculiarity of these artworks is related to their serial production by casting multiple replicas with gypsum-based plaster (also called “stucco” or “gesso”). These materials have been widely used for moulding and casting techniques which had great impulsion in Renaissance Florence2.

  • 3 Gariani et al., 2018, 2019; Beaugnon et al., 2019.

2To shed light on these artworks, previous works have attempted to find attribution criteria based on microstructural and compositional parameters3. This contribution pre-sents the work done regarding the study of manufacturing techniques and serial casting. A brief review of the casting techniques and the production chain possibly used for Florentine reliefs is essential to better understand both the issues to solve and the adopted approach. Considering Florentine devotional reliefs, the supposed phases of the serial casting process are summarized below (fig. 1).

Fig. 1. Scheme summarizing the supposed phases of the “chaîne opératoire” used for the serial production of painted Madonne di gesso

Fig. 1. Scheme summarizing the supposed phases of the “chaîne opératoire” used for the serial production of painted Madonne di gesso

© G. Gariani.

  • 4 Mozzati, 2013; Gentilini, 2012; Klapisch-Zuber, 2013.
  • 5 Gentilini, 2008; Guillet, 2017.
  • 6 See article by J. Vatelot in this volume.
  • 7 Biringuccio, 1540 (ed. 1990); Vasari 1550 (ed. 2008), Cat. Exhib. Cento, 2007; Lebrun et al., 1887; (...)
  • 8 Cennini, 1960; Carradori, 1802; Kuban, 2014; Kumar, Sisi, 1986.
  • 9 Gentilini, 2012, 2013.

3In some cases, models from the Masters, sculpted in marble, ornamental stones or metal alloys, served as prototypes4 for the serial casting of fictile workshop replicas. When the Master’s marble “progenitor” is unknown, hypotheses designate as the prototype and starting point of a series, either a prime terracotta relief or another stucco relief considered closer to the style/hand of the Master5. This prime model was then supposedly moulded. A reverse image or initial mould was then obtained and subsequently used for multiple casting of direct replicas6. The vast majority of the Florentine stucco reliefs were probably cast with plaster moulds (also called piece moulds), but other techniques were known7. For the casting of replicas, the mould was covered with a demoulding agent to favor the final release. A first thin imprint or moulding layer was applied by hand or brushed over the mould to take the imprint of all the details8. Afterwards, one or more layer/s of coarser plaster were poured to fill the mould. When the cast was partially set but not completely dry, the replicas were removed from the mould. While the plaster is still not completely set the relief could be retouched with a stick and stirring tools to reshape or correct imperfections or details. Decoration could be modified or patched by adding fresh plaster. After the complete drying, the cast can finally be painted and decorated9.

  • 10 Guillet, 2017.

4The mould was usually designed to be used repeatedly. Hence after intensive use, moulds would eventually lose their accuracy or suffer damage. Therefore, according to the needs, a new mould was created using a pristine prototype. Another option is that more moulds circulated and were available in the workshops at the same time, depending on the format or quality of the production10.

5The larger the number of replicas obtained with a mould, the higher the possibility that the latter replicas obtained could present extra flaws, modifications, or shape differences in comparison with the first ones. For instance, if a mould stayed in use for a long time, cracks, marks, or missing parts could have appeared in the obtained replicas. The details in the imprinting and the undercut parts could also end up progressively losing sharpness and accuracy. All these options only give a glimpse of how complex and challenging it is to study all the aspects related to the manufacturing techniques of multiple replicas.

  • 11 Gentilini, 2012.
  • 12 Cat. Exhib. Genève, 1997.
  • 13 Dubois, 1950; Frederiksen, 2010.

6Following the conventional art historical approach, the classification of typological series of casts is based solely on stylistic evaluation. Sometimes a qualitative estimation of the artwork’s overall dimensions can be taken into account to deduce the relationship to an original or common moulds of which no examples remain11. This conventional method of documentation does not allow the execution of an actual three-dimensional comparison. Moreover, the overall dimensions are not necessarily accurate nor conclusive to determine a direct link between two artworks. Length and height may vary due to modifications made to adapt the plaster cast to different formats or frames, or even due to deformation during demoulding or restoration of the cast. The study and comparison of all the external features are crucial: dimensions, but also volume, surface shapes, and topography, are related to the casting process and, on the other hand, on the modifications made with stirring tools or freehand modeling to retouch or add details and decorations to the plaster. These features, examined on a series of casts of the same type, can reveal information directly related to their production. It can be assumed that a replica obtained directly from a specific prototype will be closer to the latter in terms of dimensions, surface features, and shape, rather than a workshop replica, or a late copy12, obtained by using a different mould13.

Research aims

  • 14 Remondino et al., 2006, 2009; Akca et al., 2006.
  • 15 Tolksdorf et al., 2017; de Beenhouwer, 2008; Frischer, 2014; Lu et al., 2013.

7Finding similarities or differences between artworks of the same type is essential to trace a relief or a series back to a shared model/prototype. It also could be crucial to spot replicas more likely to be part of the same casting series and, as such, produced within the same workshop. An approach based on 3D digitization was adopted to compare artworks obtained through serial manufacturing. A structured light 3D scanner was tested as a simple and effective way to obtain high resolution 3D models of artworks’ entire external volume (table 1). Once digitized, surface features and topography from macro to micro (i.e. cm to mm-µm) can be studied and measured. Moreover, artworks of the same typo-logy can be systematically compared through a digital overlap of their 3D models14. It provides information on their diffe-rences in shapes and topography15 ultimately related to their level of matching. This protocol can be applied regardless of the painted decoration, which can bias or mislead the comparison of relief features and the artwork location. Through this approach, the aim was to bring new elements for the pressing questions of curators and potentially reveal:

  • divergence/matching from a supposed/unknown common model or prototype; 

  • replicas belonging to the same casting series, through the use of a shared mould, for instance;

  • changes from the original prototype, due to mass casting or differences in the customization of the decoration process.

Table 1. Set-up: scanner used and technical specifications

Table 1. Set-up: scanner used and technical specifications

8Ultimately, testing this approach could help in the attribution and classification of serially casted reliefs. The 3D models created will also serve as a basis for future studies, restoration and valorization of these long-neglected artworks.

3D scanning system and scanning workflow

  • 16 Boardman, Bryan, 2018; Jecić, Drvar, 2003; Guarato et al., 2016.
  • 17 Beraldin, 2004; Rahaman, Champion, 2019.
  • 18 Eiríksson et al., 2016.
  • 19 Pavlidis et al., 2007; Georgopoulos et al., 2010; Mara, Portl, 2013; Serna et al., 2015; Menna et a (...)
  • 20 Bénière et al., 2013, Burke et al., 2002; Malapelle et al., 2017; Zhang, 2018.
  • 21 Tolksdorf et al., 2017; de Beenhouwer, 2008; Frischer, 2014; Lu et al., 2013; Brenckmann et al., 20 (...)

9For this specific study, the main constraints were the limited time frame available, the need for an easy and ready-to-use method which is easily transportable between the different museums involved, not forgetting a high level of reproducibility and accuracy performance in measurement. Several techniques were considered as viable choices, each of them with pros and cons. Lasergrammetry16 was set apart due to possible complications related to the material’s color and aspect (polychromies) and the presence of gilding or other shiny decorations. Photogrammetry, a well-known and widespread method, both easy to implement and transportable17, could have been affected by the color texture. In addition, the level of spatial and depth resolution required to identify the details or topographical differences associated with the manufacture of these specific objects was yet to be known. In sum, to be able to take into account macro- and micro-topographic differences (down to the µm scale18), structured light was considered the most appropriate choice for this pilot study. This method is highly versatile and is increasingly applied in the field of cultural heritage19. It is also well established and used in the industrial sector for reverse engineering and quality control processes of objects produced in series from a prototype20. The principle consists of scanning an object to obtain a metrologically accurate 3D model with a resolution that can vary according to the instrument employed. Based on digital scan data, it is possible to perform a qualitative and quantitative comparison of shapes and topography. For example, two objects with similar volumes can be compared through a digital overlay of their 3D models, which provides quantified information about their correspondence. The idea was to use this well-established working method to study the serial casting of Renaissance plaster reliefs. Similar earlier works are reported on coins, figurines, and vases21.

10Two commercial structured light 3D scanning systems by AICON were used (table 1). The two instruments used can be considered equivalent and producing comparable results. The SmartScan uses the blue light fringe projection scanning method. A central projector emits different patterns of fringes on the object. Two 5 megapixel CCD monochrome cameras positioned on either side of the projector allow fast data acquisition with ultra-low noise levels and a high signal-to-noise ratio. The scanner combines three triangulation angles (10°, 20°, 30°) in each scan for optimal object coverage. This makes possible to reach narrow cavities, reduce the number of obstructed areas and ultimately capture the complete shape and geometry of the object from different angles. Given the size and shape of the objects under study, it was the scanner that was moved, avoiding unnecessary manipulation of the artwork. Another option would have been to put the sample artwork on a turning table.

  • 22 Jaquelin, 2017.
  • 23 Gariani, 2019.
  • 24 Rieke-Zapp, Trinkl, 2017.
  • 25 See article by A.-S. Le Hô et al. in this volume.
  • 26 Cignoni et al., 2008; Callieri et al., 2003; Dipanda, Woo, 2005; Scopigno et al., 2011; Serna et al (...)

11With the multidisciplinary participation of curators and conservators (plaster and sculpture experts), added to our knowledge in the field of 3D, the method was evaluated on a set of reduced-scale replicas created with controlled manufacturing processes compatible with that used in the Renaissance22. This allowed the development of an optimized workflow for in situ scanning, as described in another work23. Measurements were carried out as far as possible in darkness to avoid light interference and preceded by a calibration procedure requiring the acquisition of specific patterns on reference plates. At the end of the scanning procedure, the scans are cleared of any remaining artifacts. Then, their alignment is optimized, and they are merged with the proprietary software OptoCat to obtain the final 3D model24. In these first steps, we decided not to use an automatic “hole closing” function or automatic mesh smoothing to allow surface interpretation. We did not consider the color texture since the main objective was to study the topographical characteristics related to the moulding or manual finishing processes. The scanners automatically took a black and white capture. Other works take into account imaging analysis on polychromies25. The average time for a complete acquisition can vary depending on the conditions and the artwork; in general, it took between 1 and 4 hours to digitize a single artwork. These choices were mainly due to time constraints facilitating and accelerating information’s circulation with the curators and art historians involved in the project. Many other commercial scanners and open source software are available and allow advanced and complex data handling and treatments26.

12The resulting 3D meshes have been extremely useful for art historians: an in-depth examination of the features and details of the reliefs has been carried out by tilting and rotating them, changing the light to appreciate the relief’s features, and comparing them. The 3D meshes were aligned two by two using the same specific points. Then the positions were refined by OptoCat before launching a comparison following the vertex normals. The results were displayed by false-color topographic maps highlighting the areas of greater or lesser volumetric divergence with the relief taken as reference.

Selected artworks

  • 27 Cat. Exhib. Cento, 2007; Cat. Exhib. Paris-Florence-Washington, 2006; Guillet, 2017.
  • 28 See articles by G. Gariani, F. Beaugnon et al., A. Aksamija et al. and A.-S. Le Hô et al. in this v (...)
  • 29 See article by J. Vatelot in this volume.

13Three representative groups of low relief stucco casts after types or models of Donatello and/or followers (The Nativity), Desiderio da Settignano (Madonna with Child, type Turin), and Antonio Rossellino (Madonna with Child, type Candelabra), for a total of 12 artworks, were selected to carry out the in situ digitization campaign. Number of fundamental articles27 propose exhaustive studies and inventories of several of these typological models. The artworks are kept in six international museums: the Louvre Museum in Paris, the Museum of Fine Arts in Strasbourg, the Museum of Fine Arts in Lyon, and the Bargello Museum and Stefano Bardini Museum in Florence. These three models were ideal test subjects to assess the advantages and limitations of the approach. Being amongst the best known and most diffused, more than one example of each was available in the corpus studied for previous phases of this project28. They have different average sizes (see tables 2, 3, 4) and various levels of surface and shape complexity. The Madonna with Child (type Turin) is smaller and has a very low relief, The Nativity has larger dimensions, and the relief and modeled figures are more complex and with many undercuts. The Madonna with Child, called Madonna of the Candelabra, presents intermediate typological characteristics. It is important to mention that the overall thickness of the artworks has not been taken into account. Firstly, some of them are framed and it was neither possible nor safe to remove the frames. Second, backs of the reliefs are not necessarily pertinent; they could have been subject to removal/addition of materials or simply presenting variable thickness due to the original format or framing29. The three groups of works are reported here with information either on the assumed relationship to known master marbles (potentially causing the spread of a given type) or on known models in other plastic materials (e.g. terracotta), which might have served as a prototype/inspiration for the stucco casts.

The Nativity (after Donatello and/or followers)

14The terracotta displayed in the Bardini Museum (inv. 1200) is unanimously considered to be a 15th century original. This example has therefore been taken as a reference when trying to establish correlations with the other three stucco works (table 2).

Table 2. Reliefs studied belonging to type Nativity

Table 2. Reliefs studied belonging to type Nativity

Virgin and Child, type Turin (after Desiderio da Settignano)

15According to specialists, the model at the origin of this typology is a marble currently in the Galleria Sabauda collection in Turin, Italy. At least 20 reliefs of this model are known worldwide (table 3).

Table 3. Reliefs studied belonging to type Turin

Table 3. Reliefs studied belonging to type Turin

Madonna of the Candelabra (after Antonio Rossellino)

  • 30 Guillet, 2017.

16This composition, with over 80 known examples (table 4), is considered one of the most popular. It has been widely studied and reported in the literature. Some scholars have argued that a lost or destroyed marble may have been the source of this composition. Others suggest that the series may have been inspired by/realized from a terracotta relief now in the Victoria and Albert Museum30.

Table 4. Reliefs studied belonging to type Candelabra

Table 4. Reliefs studied belonging to type Candelabra

Results and discussion

17The adopted workflow allowed to obtain the topography of the 12 objects’ entire and complex surface. The 3D display can provide several advantages in the visualization of relief details which have better visibility and readability because of the abstraction from different textures, gilding, and polychromies that might mask them. Thanks to the available 3D models, the comparison between various works of art is easy and simultaneous. In addition, three-dimensional visualization allows examining angles and points of view that are not accessible when artworks are in an exhibition or on catalog images. This way, the comparison among several reliefs is objectively based on features related to the shapes of the reliefs, surface deformations due to demoulding or restorations, and smaller details in the finishing of the main figures of the composition.

18As a second step, mesh comparisons highlighted by false-color mapping were obtained between casts of the same type. As a general note for their interpretation, green areas indicate a high degree of matching (i.e. lower divergence in height) with the relief used as a reference for the comparison. Moving towards red and purple, the divergences in height respectively increase or decrease. In other words, the relief is either higher (red) or lower (purple) than the height of the one used as reference; therefore, it presents a lower level of local shape matching in those areas.

The Nativity

19When examining the 3D models simultaneously (fig. 2), the first observable differences between the four casts are related to the shape of the rounded upper part of the reliefs and the lateral borders, which appear more or less square from one artwork to another. This may be mostly due to the change that occurred during demoulding or the adaptation of the moulding to their various frames.

Fig. 2. 3D models obtained on four different examples of Nativity after Donatello and B. Bellano. Grids on the background provide metric references (in mm). White arrows indicate the main differences observable via qualitative examination

Fig. 2. 3D models obtained on four different examples of Nativity after Donatello and B. Bellano. Grids on the background provide metric references (in mm). White arrows indicate the main differences observable via qualitative examination

© C2RMF/G. Gariani, C. Hochart, N. Mélard/Imaging group.

20Concerning relief’s details, at first sight, the stucco inv. 1140 stands out for the absence of both the floral wreath on the ox forehead, the mane of the donkey, and the folded crest of the veil on the top of the Virgin’s head. These differences, being essentially decorative details, may not have a fundamental significance. In addition, the level of “finesse” of the modeling, the sharpness of detail’s imprinting, especially in the main figures’ faces, are more pronounced and refined in the terracotta (inv. 1200) and decreasingly in stucco RF 1191 and D 488. On the other hand, inv. 1140’s modeling seems rougher with less attention applied to the imprinted details and some of the jutting parts (e.g. St. Joseph and the Virgin noses are roughly squared and pointy).

  • 31 Gariani, 2019.
  • 32 Cat. Exhib. Cento, 2007.

21To quantify the overall differences and visually localize their extension, a 3D shape comparison, based on the artworks’ scanned surface geometry, was performed31. Based on the literature, the terracotta example (inv. 1200) from the Bardini Museum could have been at the origin of a casting series created using this very example as a prototype32. Therefore, inv. 1200 was used as a reference to be compared with the other three reliefs of the same typology. Results are shown in a false-color mapping (fig. 3).

Fig. 3. Topographic false color cartography obtained through the mesh comparison of the 3D models using as reference Nativity inv. 1200 (terracotta, Bardini Museum, Florence). Green areas indicate higher matching between the reliefs, whilst in areas going towards purple and red the matching worsen. Comparison between: a] Inv. 1140 and inv. 1200 b) RF 1191 and inv. 1200; c] D 488 and inv. 1200

Fig. 3. Topographic false color cartography obtained through the mesh comparison of the 3D models using as reference Nativity inv. 1200 (terracotta, Bardini Museum, Florence). Green areas indicate higher matching between the reliefs, whilst in areas going towards purple and red the matching worsen. Comparison between: a] Inv. 1140 and inv. 1200 b) RF 1191 and inv. 1200; c] D 488 and inv. 1200

© C2RMF/G. Gariani/Imaging group.

22The comparison between inv. 1200 and inv. 1140 (fig. 3 a) shows an average difference in relief height of ±3 mm with some parts of inv. 1140 showing a divergence of up to 6 mm. The highest differences are located mainly in the background and the parts with higher reliefs.

23The situation changes considering the level of matching between inv. 1200 and RF 1191 (fig. 3 b). The average divergence in height between these two reliefs is about ±2.3 mm. A similar situation can be observed between D 488 and inv. 1200 with a difference of around ±2.4 mm (fig. 3 c). Comparing D 488 with RF 1191, there is significant accordance (±2.4 mm) too, even if the dimensional matching between these two examples seems to cover a smaller portion of the relief surface.

24A clear and consistent correlation and good shape matching relate to three artworks, i.e. inv. 1200, RF 1191 and D 488. Therefore, we can assume that RF 1191, and plausibly also D 488, could have been made as a direct replica of inv. 1200, or at least three artworks share a common prototype from which they were all reproduced. The other relief of the Bardini Museum (inv. 1140) appears systematically different from the others. Two main explanations are possible. This cast could also be a direct replica of inv. 1200; in this case, more pronounced differences may be due to deformations occurring during the demoulding with consequent loss of details and retouches of the more jutting parts (e.g noses, Virgin’s veil). Otherwise, the cast might be a workshop replica obtained with another mould and belonging to a different casting series. The analyses of the materials would be required to have more confirmations.

Virgin and Child (type Turin)

  • 33 Cat. Exhib. Florence-Paris, 2013, p. 426-429, VIII.2-3.

25Looking at the 3D models (fig. 4), the Louvre cast (RF 897) shows a higher level of reproduction and precision used for the cast details (e.g. the drapery, hair and face of the Virgin are very refined). The other two examples (inv. 1194 and D 489) show more pronounced differences in the overall dimensions and additional details in the decorations. Moreover, the details of the imprinting seem more flattened and less sharp. It might also be due to a better perception and detection of these details because of the absence of layers of polychromy on RF 897’s surface. This is consistent with the literature33 identifying the relief kept at the Louvre (RF 897) as a direct replica of the marble attributed to Desiderio and exhibited at the Galleria Sabauda in Turin (Italy). For this reason, RF 897 was used as a reference for the mesh comparison with the other two low reliefs (fig. 5).

Fig. 4. 3D models of the three digitized examples of the type Turin. White arrows indicate some of the main differences, related to the decoration of the background and to the Virgin’s veil which is almost not visible in inv. 1194 (c). Also, the level of sharpness in relief’s details is higher in RF 897, and decrease in (a), D 489 (b), to inv. 1194 (c)

Fig. 4. 3D models of the three digitized examples of the type Turin. White arrows indicate some of the main differences, related to the decoration of the background and to the Virgin’s veil which is almost not visible in inv. 1194 (c). Also, the level of sharpness in relief’s details is higher in RF 897, and decrease in (a), D 489 (b), to inv. 1194 (c)

© C2RMF/G. Gariani, C. Hochart/Imaging group.

Fig. 5. 3D comparison between a] Inv. 1194 and RF 897 used as reference; b] D 489 and RF 897 used as reference c) D 489 and inv. 1194 used as reference. Green areas indicate lower divergence (i.e. higher matching) from the model used as a reference

Fig. 5. 3D comparison between a] Inv. 1194 and RF 897 used as reference; b] D 489 and RF 897 used as reference c) D 489 and inv. 1194 used as reference. Green areas indicate lower divergence (i.e. higher matching) from the model used as a reference

© C2RMF/G. Gariani/Imaging group.

26From the comparison between inv. 1194 and RF 897 (fig. 5 a), it can be seen how the Bardini relief (inv. 1194) presents a main central area of divergence (visible in blue) where the surface is lower by about 3 to 6 mm. This difference is due to the absence of the veil around the neck and on the chest of the Virgin. Otherwise, the average divergence between the objects is +2.3 mm. Relief D 489 has an average surface matching of + 2.0 mm with RF 897 (fig. 5 b). Considering the comparison between D 489 and inv. 1194 (fig. 5 c), one can see how the differences between the two are slightly lower (+1.8 mm). These two reliefs, therefore, present a closer matching than with artwork RF 897. This matching between D 489 and inv. 1194, and the fact that they have different backgrounds and decorations, might confirm their classification as part of a series of workshop replicas (maybe obtained with different moulds) or maybe as two replicas customized according to buyers’ requests.

Virgin and Child (type Candelabra)

  • 34 Guillet, 2017.
  • 35 Le Hô et al., 2018. See article by A.-S. Le Hô et al. in this volume.

27The study of these five casts presented a more complicated case. Since the original model is unknown, the comparisons were made by testing as possible reference three of the five models, namely inv. 422 (Bargello Museum), Campana 20 (Louvre Museum), and DS 534 (Écouen Renaissance Museum). The MBA 587 was excluded from the selection even though recent studies have suggested a close relationship between the stucco cast and the Victoria and Albert Museum terracotta, thus regarded as the model upon which this type derives34. The choice, advised by curators, was motivated by the known history and provenance of the artworks, stylistic examinations, and the presence of polychromy considered as original from the 15th century and subject to fewer repainting interventions35.

28By a close 3D examination visible from the close-up images (fig. 6 a), differences in the Virgin’s head are visible in the shape of the nose, the hair, the presence/absence of a headband on the forehead, and of the top part of the Virgin’s veil.

Fig. 6. Close up on a) the Virgin’s face, and on b) the Child’s head of the five Madonna of the Candelabra digitized. White arrows indicate some of the main differences observable

Fig. 6. Close up on a) the Virgin’s face, and on b) the Child’s head of the five Madonna of the Candelabra digitized. White arrows indicate some of the main differences observable

© C2RMF/G. Gariani/Imaging group.

29Some variability can also be seen by examining the child’s face (fig. 6 b) with different nose and mouth shapes. The finish and details of the curly hair progressively less defined in DS 534, inv. 422, MBA 587, Campana 20, 1937-4. The left ear is more rounded in Campana 20 and MBA 587 and more squared in the other three. The forms and decoration of the halo may also vary, with similarities observed on four reliefs and a different, flatter decoration for inv. 422.

30The comparisons displayed by the false-color mapping allow the height differences of the reliefs to be better quantified. Each relief was alternatively compared with the three considered as possible reference for the casting series (fig. 7). Figure 8 summarizes schematically the values of average divergence resulting from each comparison.

Fig. 7. Topographic false color comparison of the 3D models of the five Madonna of the Candelabra stucco casts digitized. Inv. 422 (Bargello Museum), Campana 20 (Louvre Museum), and DS 534 (Écouen Renaissance Museum) were used as reference for comparisons. Moving away toward purple and red the level of matching worsen

Fig. 7. Topographic false color comparison of the 3D models of the five Madonna of the Candelabra stucco casts digitized. Inv. 422 (Bargello Museum), Campana 20 (Louvre Museum), and DS 534 (Écouen Renaissance Museum) were used as reference for comparisons. Moving away toward purple and red the level of matching worsen

© C2RMF/G. Gariani/Imaging group.

Fig. 8. Average divergence obtained by mesh comparisons on the 3D models of the Madonna of the Candelabra studied; the ones used as reference are respectively indicated as title of each of the three histograms

Fig. 8. Average divergence obtained by mesh comparisons on the 3D models of the Madonna of the Candelabra studied; the ones used as reference are respectively indicated as title of each of the three histograms

© C2RMF/G. Gariani.

31Results (fig. 7) often reveal background protrusion or convexity possibly related to a certain degree of plastic deformation experienced by the casts during their demoulding and maybe in some cases (MBA 587 and DS 534) during restoration. Comparisons can also be more or less affected by differences in dimensions, framing, structural problems, and fractures. In particular, DS 534 tends to have higher divergence values (i.e. a lower shape match) with the other reliefs. The other four appear to be closer to one another in terms of differences.

Synthesis and conclusions

32This work enlightened the results obtained by applying a structured light 3D scanner to study Renaissance devotional reliefs’ serial casting. The employed protocol provided art historians and curators with an additional tool to improve the classification based on style. Thus, the volumetric and formal characteristics of these works of art are examined more closely and metrologically. First hypotheses were proposed concerning the relationship among the digitized artworks of the three studied groups of Madonne di gesso.

The Nativity (Donatello and/or followers)

  • 36 Beaugnon et al., 2019; Gariani et al., 2018 and 2019.

33A straightforward correlation has been established between three of the four examples. Reliefs RF 1191 (Louvre Museum) and D 488 (Museum of Fine Arts in Lyon) were found to be closer (better match of shape and topography) to the supposedly first terracotta model (inv. 1200) in the Bardini Museum. It is consistent with the results on the constituent materials36, which are very similar for RF 1191 and D 488, and suggests that both were probably produced in the same work-shop. It seems plausible that RF 1191 and D 488 came from direct casts of the Bardini terracotta. However, relief in stucco inv. 1140 (Bardini Museum) displayed the lowest level of shape matching. In the future, it would be interesting to analyze the constitutive materials of this example to further study its possible correlations with other elements of the series.

Virgin and Child, type Turin (Desiderio da Settignano)

  • 37 See article by G. Gariani, F. Beaugnon et al. in this volume.

34Results are consistent with most accredited art historians’ hypotheses. The features of the relief of the Louvre Museum (RF 897) most likely suggest that this cast, highly refined and detailed in its moulding features, could be a direct replica of Desiderio’s marble stored in Torino (Italy). For improved results, it would be necessary to digitize directly the latter for further comparisons. The other two reliefs examined (inv. 1194 and D 489) are very similar, although quite different from RF 897 regarding the dimensions, decorations, flattening, and smoothening of some parts of the model. However, the characterization of the materials of D 489 gave results very consistent with those of RF 89737 implying their production with the same materials. These indications could identify D 489, and plausibly inv. 1194, as workshop replicas. Perhaps they were obtained when the mould was already degraded, with a different mould, or simply decorated or personalized differently.

Virgin and Child, type Candelabra (Antonio Rossellino)

35From these first comparisons, inv. 422 or Campana 20 seem the more plausible option for a shared prototype considering the lower divergences (i.e. higher surface matching) measured with the other reliefs. Although, it is not yet possible to draw straightforward conclusions. Even if DS 534 seems less likely to have been used as a prime model for the other examined casts, it is worth noting that this relief probably suffered significant deformations, either during its demoulding or more probably due to major past restoration processes. It may have led to changes, such as in the convexity of the background, ultimately distorted the comparison between the casts. Further investigation on other examples of this type is required.

  • 38 De Luca et al., 2011; Frischer, 2014.

36In the future, other methods and software will be further evaluated to keep up with the swift progress that this field of 3D imaging and digital application in cultural heritage is experiencing. Moreover, it would be worthwhile to implement data processing and explore possibilities to automate comparisons via deep learning approaches. This point also brings our attention to data management, which is crucial when dealing with transnational projects studying serial artistic productions; complications reside in the large number and widespread distribution of these artworks in museums. It would be interesting to use a platform for specific serial artistic productions, crucial for the valorization of these lesser known though important cultural artefacts. As it has already been seen in the cultural heritage field38, 3D models could serve as a basis for semantic annotation of other data (e.g. color mapping, other imaging methods, material analysis). Furthermore, as many of these works of art have been somewhat overlooked during their history, virtual restorations or 3D printing to complete missing parts could be considered.

37In conclusion, even if the number of reliefs scanned is still limited, this was a significant starting point that will hopefully prompt the use of this approach on a larger number of reliefs and types of Renaissance Madonne di gesso. More broadly, this work could also increase the awareness of art historians and curators to the possibilities offered by 3D imaging studying serial artistic productions.

The scanning of the artworks was allowed by courtesy of Marc Bormand (musée du Louvre), Paola D’Agostino (Museo Nazionale del Bargello), Ilaria Ciseri (Museo Nazionale del Bargello), Antonella Nesi (Museo Stefano Bardini), Ludmila Virassamynaïken (musée des Beaux-Arts de Lyon), Thierry Crépin-Leblond (Musée national de la Renaissance, Écouen), and Dominique Jacquot (musée des Beaux-Arts de Strasbourg). Their support and contribution were essential for this research. The authors gratefully acknowledge Monica Galeotti from the Opificio delle Pietre Dure, the staff of all the museums, and the municipality of Florence. Léa Jaquelin, Hélène Susini and Jennifer Vatelot are gratefully acknowledged for the collaboration and the work done on reduced scale replicates and the knowledge on casting processes. Anne Bouquillon, Fabrice Goubard and Clotilde Boust are thankfully acknowledged. AICON 3D Systems GmbH part of Hexagon conferred the Bernd Breuckmann Award for 3D scanning projects in Arts and Cultural Heritage, 2017 edition. The prize was bestowed for the project: “ReMaStER3D (Renaissance Masters Stucco dEvotional Reliefs 3Dscanning), Scanning hidden secrets of a serial production from Old Masters workshops”. Dirk Rieke-Zapp and Bernd Breuckmann are thankfully acknowledged for the support provided. This work was supported by the Paris Seine Graduate School Humanities, Creation, Heritage, Investissement d’Avenir ANR-17-EURE-0021 – Foundation for Cultural Heritage Science.

Haut de page

Bibliographie

Akca D., Remondino F., Novák D., Hanusch T., Schrotter G., Gruen A., 2006, “Recording and modeling of cultural heritage objects with coded structured light projection systems”, Second international conference in remote sensing (Rome, 4-7th December 2006), p. 375-382.

Aksamija A., Nowik W., Lehuédé P., Bormand M., Bouquillon A., 2019, “Investigation of organic additives in Italian Renaissance devotion stucco reliefs from French collections”, Journal of Cultural Heritage, 39, p. 66-81.

Beaugnon F., Gariani G., Gouillart E., Bouquillon A., Bormand M., Wallez G., 2019, “Microstructure imaging of Florentine stuccoes through x-Ray tomography: a new insight on ancient plaster-making techniques”, Journal of Cultural Heritage, 40, p. 17-24.

Bénière R., Gesquière G., Le Breton F., Puech W., Subsol G., 2013, “A comprehensive process of reverse engineering from 3D meshes to CAD models”, Computer-Aided Design, 45, Elsevier, p. 1382- 1393.

Beraldin J.-A., 2004, “Integration of laser scanning and close range photogrammetry – The last decade and beyond”, International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 35 (B7), p. 972-983.

Biringuccio V., [1540], 1990, The Pirotechnia of Vannoccio Biringuccio: The Classic Sixteenth-Century Treatise on Metals and Metallurgy, Dover Publications, New York.

Boardman C., Bryan P., 2018, “3D Laser Scanning for Heritage. Advice and guidance on the use of laser scanning in archaeology and architecture”, Historic England, p. 58-60.

Breuckmann B., Karl S., Trinkl E., 2013, “Digitising Ancient Pottery. Precision in 3D”, Forum Archaeologiae, 66/III [http://farch.net].

Burke J., Bothe T., Osten W., Hess C., 2002, “Reverse engineering by fringe projection”, Interferometry XI: Applications, Proceedings of SPIE, vol. 4778, p. 312-324.

Callieri M., Cignoni P., Ganovelli F., Montani C., Pingi P., Scopigno R., 2003, “VCLab’s Tools for 3D range data processing”, VAST’03, Brighton, p. 13-22.

Carradori F., 1802, Instruzione elementare per gli studiosi della scultura, di Francesco Carradori professore di dett’arte nella R. Scuola di Firenze, Florence.

Cat. Exhib. Cento, 2007, La Madonna del presepe, da Donatello a Guercino. Una devozione antica e nuova nella terra di Cento [Exhib. Cento, Minerva Edizioni, Pinacoteca Civita, 2007, 1st December 2007-13th April 2008], G. Adani et al. (eds.), Minerva Edizioni, Argelato.

Cat. Exhib. Florence-Paris, 2013, Le Printemps de la Renaissance. La sculpture et les arts à Florence 1400-1460 [Exhib. Florence, palais Strozzi, 23th March-18th August 2013 ; Paris, musée du Louvre, 26th September 2013-6th January 2014], M. Bormand, B. Paolozzi Strozzi (eds.), Musée du Louvre Éditions, Paris/Officina Libraria, Rome.

Cat. Exhib. Genève, 1997, L’art d’imiter. Images de la Renaissance italienne au Musée d’art et d’histoire ; falsifications, manipulations, pastiches [Exhib. Genève, Musée d’art et d’histoire], Ritschard C., M. Natale (eds.), Musée d’art et d’histoire, Genève.

Cat. Exhib. Paris-Florence-Washington, 2006, Desiderio da Settignano : sculpteur de la Renaissance florentine [Exhib. Paris, musée du Louvre ; Florence, Museo Nazionale del Bargello ; Washington, National Gallery of Art, 2006-2007], B. Paolozzi Strozzi, M. Bormand, N. Penny (eds.), 5 Continents Éditions, Milan/National Gallery of Art, Washington/Musée du Louvre Éditions, Paris.

Cennini C., 1960, The Craftsman’s Handbook. “Il Libro dell’Arte”, trans. by D. V. Thompson Jr, Dover Publications, New York.

Cignoni P., Callieri M., Corsini M., Dellepiane M., Ganovelli F., Ranzuglia G., 2008, “Meshlab: an open-source mesh processing tool” , Eurographics Italian chapter conference, p. 129-136.

de Beenhouwer J., 2008, “Data management for moulded ceramics and digital image comparison. A case study of Roman terracotta figurines”, Layers of Perception. Proceedings of the 35th International Conference on Computer Applications and Quantitative Methods in Archaeology (CAA) (Berlin 2-6th April 2007; Bonn, 2008), p. 160-163.

De Luca L., Busayarat C., Stefani C., Véron P., Florenzano M., 2011, “A semantic-based platform for the digital analysis of architectural heritage”, Computers & Graphics, 35, p. 227-241.

Dipanda A., Woo S., 2005, “Efficient correspondence problem-solving in 3-D shape reconstruction using a structured light system”, Optical Engineering, 44, p. 1-14.

Dubois E., 1950, La statuaire : modelage, moulage et terre cuite ; statues, bustes, bas-reliefs, ornements d’architecture ; argile, cire, plâtre, marbre, pierre, bois, albâtre, ivoire, etc., bronze d’art, Ed. H. Delarue, Paris.

Eiríksson E. R., Wilm J., Pedersen D. B., Aanæs H., 2016, “Precision and accuracy parameters in structured light 3D scanning”, ISPRS – International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XL-5/W8, p. 7-15.

Frederiksen R., 2010, “Plaster casts in Antiquity”, in R. Frederiksen and E. Marchand (eds.), Plaster casts: making, collecting and displaying from classical Antiquity to the present (Transformationen der Antike), 18, De Gruyter, Berlin/New York, p. 13-34.

Frischer B., 2014, “3D data capture, restoration and online publication of sculpture”, in F. Remondino, S. Campana (eds.), 3D Recording and Modelling in Archaeology and Cultural Heritage, p. 137-144.

Gariani G., Lehuédé P., Leroux L., Wallez G., Goubard F., Bouquillon A., Bormand M., 2018, “First insights on the mineral composition of “stucco” devotional reliefs from Italian Renaissance Masters: investigating technological practices and raw material sourcing”, Journal of Cultural Heritage, vol. 34, p. 23-32 [doi.org/10.1016/j.culher.2018.05.003].

Gaurico P., [1504], 1969, De sculptura, A. Chastel, R. Klein (eds.), Librairie Droz, Genève-Paris.

Gentilini G., 2008, Dal rilievo alla pittura. La Madonna delle Candelabre di Antonio Rossellino, Edizioni Polistampa, Florence.

Gentilini G., 2012, “Scultura dipinta o pittura a rilievo? Riflessioni sulla policromia nel quattrocento fiorentino”, Technè, 36, p. 9-17.

Gentilini G., 2013, Scultura italiana del Rinascimento, statue e rilievi in marmo e pietra, terracotta, stucco e legno, bronzetti e sculture decorative, Giorgio Baratti Galleria, Edizioni Polistampa, Florence.

Georgopoulos A., Ioannidis C., Valanis A., 2010, “Assessing the performance of a structured light scanner”, ISPRS – International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XXXVIII-5, p. 250-255 [https://www.isprs.org/proceedings/xxxviii/part5/papers/177.pdf].

Guarato A. Z., Loja A. C., Pereira L. P., Braga S. L., Trevilato T. R. B, 2016, “Qualification of a 3D structured light sensor for a reverse engineering application”, Proceedings of SPIE 10151, Optics and Measurement International Conference, 101510C (11th November 2016)[https://doi.org/10.1117/12.2257601].

Hochart C., Lambert E., 2019, “Scanning the Celts: evaluation of 2D and 3D techniques in protohistoric archaeology”, Proceedings of SPIE, 11058, Optics for Arts, Architecture, and Archaeology VII, 110581L (22th July 2019) [doi.org/10.1117/12.2527566].

Jecić S., Drvar N., 2003, “The assessment of structured light and laser scanning methods in 3D shape measurements”, 4th International Congress of Croatian Society of Mechanics (Bizovac, Croatia, 18-20th September 2003), p. 237-244.

Johnson G. A., 1997, “Art or artefact? Madonna and Child reliefs in the early Renaissance”, in S. Currie, P. Motture (eds.), The Sculpted Object 1400-1700, Scholar Press, Aldershot, p. 1-17.

Klapisch-Zuber C., 2013, “Adorer, apprécier, afficher : les sculptures domestiques”, in Cat. Exhib. Florence-Paris, 2013, Le Printemps de la Renaissance. La sculpture et les arts à Florence 1400-1460 [Exhib. Florence, palais Strozzi ; Paris, musée du Louvre, 2013-2014], M. Bormand, B. Paolozzi Strozzi (eds.), Musée du Louvre Éditions, Paris/Officina Libraria, Rome, pages 229-236.

Kuban G. J., 2014, Making Plaster (Gypsum) Casts – Summary 1996-2014 [http://paleo.cc/casting/plaster-casting.htm].

Kumar F., Sisi C., 1986, “Un rilievo in stucco del Museo Bardini: La tecnica e il restauro”, OPD Restauro, 1, p. 10-11, Centro Di Della Edifimi SRL, Florence.

Lebrun R. F.-B., Magnier M., Robert D., de Valicourt E., 1887, Nouveau manuel complet du mouleur en plâtre, au ciment, à l’argile, à la cire, à la gélatine, Librairie encyclopédique de Roret, Paris.

Lee Palmer A., 2001, “The Walters’ “Madonna and Child”. Plaquette and private devotional art in Early Renaissance Italy”, The Journal of the Walters Art Museum, Focus on the Collections, vol. 59, p. 3-84.

Le Hô A.-S., Pingaud N., Vandenberghe Y., 2018, “Les reliefs en plâtre peint de la Renaissance italienne”, GRPA – Le plâtre et la couleur 2 – Le plâtre peint, Journée d’étude sur le Plâtre peint (23th March 2018), p. 3-6.

Lu M. et al., 2013, “Portrait sculptures of Augustus: categorization via local shape comparison”, Digital Heritage International Congress (Marseille, 28 October-1st November 2013) [dx.doi.org/10.1109/DigitalHeritage.2013.6743812].

Malapelle F., Dall’Alba D., Dalla Fontana D., Dall’Alba I., Fiorini P., Muradore R., 2017, “Cost effective quality assessment in industrial parts manufacturing via optical acquisition”, Procedia Manufacturing, vol. 11, p. 1207-1214.

Mara H., Portl J., 2013, “Acquisition and documentation of vessels using high-resolution 3D-scanners (Hrsg.)”, in E. Trinkl (ed.), Interdisziplinäre Dokumentations- und Visualisierungsmethoden, CVA Österreich Beiheft 1, Verlag Der Österreichischen Akademie der Wissenschaften, Vienna, p. 25-40.

Menna F., Nocerino E., Remondino F., Dellepiane M., Callieri M., Scopigno R., 2016, “3D Digitization of an Heritage Masterpiece – a critical analysis of quality assessment”, ISPRS – International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLI-B5 [10.5194/isprsarchives-XLI-B5-675-2016].

Mozzati T., 2013, “À la manière des Florentins, Diffusion et tendances du goût à l’époque de la Renaissance italienne”, in Cat. Exhib. Florence-Paris, 2013, Le Printemps de la Renaissance. La sculpture et les arts à Florence 1400-1460 [Exhib. Florence, palais Strozzi ; Paris, musée du Louvre, 2013-2014], M. Bormand et B. Paolozzi Strozzi (eds.), Musée du Louvre Éditions, Paris/Officina Libraria, Rome, p. 181-187.

Müller A., 1997, “Description et analyse des productions moulées : proposition de méthode”, in A. Müller (ed.), Le moulage en terre cuite dans l’Antiquité : création et production dérivée, fabrication et diffusion, Actes du 18e Colloque du Centre de Recherches Archéologiques Lille III (7-8 déc. 1995), p. 437-463, Presses universitaires du Septentrion, Lille.

Pavlidis G., Koutsoudis A., Arnaoutoglou F., Tsioukas V., Chamzas C., 2007, “Methods for 3D digitization of Cultural Heritage”, Journal of Cultural Heritage, vol. 8, p. 93-98.

Payne E., 2016, “Comparative 3D scanning of historical casts : the Parthenon casts at the British Museum”, in C. Haak, M. Helfrich (eds.), Casting. A way to embrace the digital age in analogue fashion? A symposium on the Gipsformerei of the Staatlichen Museen zu Berlin, Universitätsbibliothek Heidelberg, arthistoricum.net [https://doi.org/10.11588/arthistoricum.95.114].

Rahaman H., Champion E., 2019, “To 3D or Not 3D: Choosing a photogrammetry workflow for Cultural Heritage Groups”, Heritage, 2, p. 1835-1851 [doi.org/10.3390/heritage2030112].

Remondino F., El-Hakim S., 2006, “Image-based 3D modelling: a review”, Photogrammetric Record, 21 (115), p. 269-291.

Remondino F., Girardi S., Rizzi A., Gonzo L., 2009, “3D modelling of Complex and Detailed Cultural Heritage using Multi-Resolution Data”, Journal of Computing and Cultural Heritage, 2 [doi: 10.1145/1551676.1551678].

Rieke-Zapp D. H., Trinkl E., 2017, “Face to face – close range inspection of head vases”, ISPRS – International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-2/W5, p. 601-604 [doi.org/10.5194/isprs-archives-XLII-2-W5-601-2017].

Scopigno R., Callieri M., Cignoni P., Corsini M., Dellepiane M., Ponchio F., Ranzugli G., 2011, “3D models for Cultural Heritage: beyond plain visualization”, IEEE Computer, vol. 44, n° 7, p. 48-55.

Serna C. G., Pillay R., Trémeau A., 2015, “Data fusion of objects using techniques such as laser scanning, structured light and photogrammetry for cultural heritage applications”, in A. Trémeau, R. Schettini, S. Tominaga (eds.), Computational Color Imaging. CCIW 2015. Lecture Notes in Computer Science, vol. 9016, p. 208-224.

Tolksdorf J. F., Elburg R., Reuter T., 2017, “Can 3D scanning of countermarks on Roman coins help to reconstruct the movement of Varus and his legion”, Journal of Archaeological Science: Reports, 11, p. 400-410.

Trinkl E., Rieke-Zapp D., Homer L., 2019, “Face to face – Considering the moulding of Attic head vases reconsidering Beazley’s groups by quantitative analysis”, Journal of Archaeological Science: Reports, 21, p. 1019-1024.

Vasari G., [1550], 2008, The Lives of the Artists, trans. by J. C. and P. Bondanella, Oxford University Press, Oxford.

Zhang S., 2018, “High-speed 3D shape measurement with structured light methods: A review”, Optics and Lasers in Engineering, 106, p. 119-131 [doi.org/10.1016/j.optlaseng.2018.02.017].

Unpublished documents

Gariani G., 2019, Unravelling materials, provenance, and serial manufacturing of the Madonne di gesso: A first technical study on stucco devotional reliefs from Italian Renaissance Masters. PhD Thesis, under the supervision of F. Goubard et A. Bouquillon, École doctorale Sciences et Ingénierie, université de Cergy-Pontoise, Val d’Oise.

Guillet S., 2017, Les reliefs en stuc polychromé de La Vierge et l’Enfant par Antonio Rossellino et son atelier. Master 2 in art history, under the supervision of M. Bormand et P. Sénéchal, École du Louvre, Paris.

Jaquelin L., 2017, Techniques d’estampages et de moulage à la Renaissance Italienne, Etude Historique et expérimentale. Master 2 Conservation-Restauration des Biens Culturels, under the supervision of A. Bouquillon and H. Susini, Université Paris 1 Panthéon Sorbonne.

Web references

AICON 3D Systems, 2016. GmbH, OptoCat User Manual.

Haut de page

Notes

1 Gentilini, 2012; Cat. Exhib. Florence-Paris, 2013, p. 426-429, VIII.2-3.

2 Vasari, 1550 (ed. 2008); Gaurico, 1504 (ed. 1969); Biringuccio, 1540 (ed. 1990); Lebrun et al., 1887; Frederiksen, 2010; Johnson, 1997; Lee Palmer, 2001.

3 Gariani et al., 2018, 2019; Beaugnon et al., 2019.

4 Mozzati, 2013; Gentilini, 2012; Klapisch-Zuber, 2013.

5 Gentilini, 2008; Guillet, 2017.

6 See article by J. Vatelot in this volume.

7 Biringuccio, 1540 (ed. 1990); Vasari 1550 (ed. 2008), Cat. Exhib. Cento, 2007; Lebrun et al., 1887; Müller, 1997.

8 Cennini, 1960; Carradori, 1802; Kuban, 2014; Kumar, Sisi, 1986.

9 Gentilini, 2012, 2013.

10 Guillet, 2017.

11 Gentilini, 2012.

12 Cat. Exhib. Genève, 1997.

13 Dubois, 1950; Frederiksen, 2010.

14 Remondino et al., 2006, 2009; Akca et al., 2006.

15 Tolksdorf et al., 2017; de Beenhouwer, 2008; Frischer, 2014; Lu et al., 2013.

16 Boardman, Bryan, 2018; Jecić, Drvar, 2003; Guarato et al., 2016.

17 Beraldin, 2004; Rahaman, Champion, 2019.

18 Eiríksson et al., 2016.

19 Pavlidis et al., 2007; Georgopoulos et al., 2010; Mara, Portl, 2013; Serna et al., 2015; Menna et al., 2016.

20 Bénière et al., 2013, Burke et al., 2002; Malapelle et al., 2017; Zhang, 2018.

21 Tolksdorf et al., 2017; de Beenhouwer, 2008; Frischer, 2014; Lu et al., 2013; Brenckmann et al., 2013; Payne, 2016; Trinkl et al., 2019.

22 Jaquelin, 2017.

23 Gariani, 2019.

24 Rieke-Zapp, Trinkl, 2017.

25 See article by A.-S. Le Hô et al. in this volume.

26 Cignoni et al., 2008; Callieri et al., 2003; Dipanda, Woo, 2005; Scopigno et al., 2011; Serna et al., 2015.

27 Cat. Exhib. Cento, 2007; Cat. Exhib. Paris-Florence-Washington, 2006; Guillet, 2017.

28 See articles by G. Gariani, F. Beaugnon et al., A. Aksamija et al. and A.-S. Le Hô et al. in this volume.

29 See article by J. Vatelot in this volume.

30 Guillet, 2017.

31 Gariani, 2019.

32 Cat. Exhib. Cento, 2007.

33 Cat. Exhib. Florence-Paris, 2013, p. 426-429, VIII.2-3.

34 Guillet, 2017.

35 Le Hô et al., 2018. See article by A.-S. Le Hô et al. in this volume.

36 Beaugnon et al., 2019; Gariani et al., 2018 and 2019.

37 See article by G. Gariani, F. Beaugnon et al. in this volume.

38 De Luca et al., 2011; Frischer, 2014.

Haut de page

Table des illustrations

Titre Fig. 1. Scheme summarizing the supposed phases of the “chaîne opératoire” used for the serial production of painted Madonne di gesso
Crédits © G. Gariani.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-1.jpg
Fichier image/jpeg, 284k
Titre Table 1. Set-up: scanner used and technical specifications
URL http://journals.openedition.org/techne/docannexe/image/9127/img-2.png
Fichier image/png, 38k
Titre Table 2. Reliefs studied belonging to type Nativity
URL http://journals.openedition.org/techne/docannexe/image/9127/img-3.jpg
Fichier image/jpeg, 200k
Titre Table 3. Reliefs studied belonging to type Turin
URL http://journals.openedition.org/techne/docannexe/image/9127/img-4.jpg
Fichier image/jpeg, 184k
Titre Table 4. Reliefs studied belonging to type Candelabra
URL http://journals.openedition.org/techne/docannexe/image/9127/img-5.jpg
Fichier image/jpeg, 412k
Titre Fig. 2. 3D models obtained on four different examples of Nativity after Donatello and B. Bellano. Grids on the background provide metric references (in mm). White arrows indicate the main differences observable via qualitative examination
Crédits © C2RMF/G. Gariani, C. Hochart, N. Mélard/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-6.jpg
Fichier image/jpeg, 432k
Titre Fig. 3. Topographic false color cartography obtained through the mesh comparison of the 3D models using as reference Nativity inv. 1200 (terracotta, Bardini Museum, Florence). Green areas indicate higher matching between the reliefs, whilst in areas going towards purple and red the matching worsen. Comparison between: a] Inv. 1140 and inv. 1200 b) RF 1191 and inv. 1200; c] D 488 and inv. 1200
Crédits © C2RMF/G. Gariani/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-7.jpg
Fichier image/jpeg, 1,2M
Titre Fig. 4. 3D models of the three digitized examples of the type Turin. White arrows indicate some of the main differences, related to the decoration of the background and to the Virgin’s veil which is almost not visible in inv. 1194 (c). Also, the level of sharpness in relief’s details is higher in RF 897, and decrease in (a), D 489 (b), to inv. 1194 (c)
Crédits © C2RMF/G. Gariani, C. Hochart/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-8.jpg
Fichier image/jpeg, 284k
Titre Fig. 5. 3D comparison between a] Inv. 1194 and RF 897 used as reference; b] D 489 and RF 897 used as reference c) D 489 and inv. 1194 used as reference. Green areas indicate lower divergence (i.e. higher matching) from the model used as a reference
Crédits © C2RMF/G. Gariani/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-9.jpg
Fichier image/jpeg, 304k
Titre Fig. 6. Close up on a) the Virgin’s face, and on b) the Child’s head of the five Madonna of the Candelabra digitized. White arrows indicate some of the main differences observable
Crédits © C2RMF/G. Gariani/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-10.jpg
Fichier image/jpeg, 324k
Titre Fig. 7. Topographic false color comparison of the 3D models of the five Madonna of the Candelabra stucco casts digitized. Inv. 422 (Bargello Museum), Campana 20 (Louvre Museum), and DS 534 (Écouen Renaissance Museum) were used as reference for comparisons. Moving away toward purple and red the level of matching worsen
Crédits © C2RMF/G. Gariani/Imaging group.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-11.jpg
Fichier image/jpeg, 640k
Titre Fig. 8. Average divergence obtained by mesh comparisons on the 3D models of the Madonna of the Candelabra studied; the ones used as reference are respectively indicated as title of each of the three histograms
Crédits © C2RMF/G. Gariani.
URL http://journals.openedition.org/techne/docannexe/image/9127/img-12.jpg
Fichier image/jpeg, 136k
Haut de page

Pour citer cet article

Référence papier

Gianluca Gariani, Charlotte Hochart et Nicolas Mélard, « Renaissance plaster casts: first insights by means of Structured Light 3D Scanner »Technè, 51 | 2021, 105-117.

Référence électronique

Gianluca Gariani, Charlotte Hochart et Nicolas Mélard, « Renaissance plaster casts: first insights by means of Structured Light 3D Scanner »Technè [En ligne], 51 | 2021, mis en ligne le 15 décembre 2022, consulté le 17 janvier 2026. URL : http://journals.openedition.org/techne/9127 ; DOI : https://doi.org/10.4000/techne.9127

Haut de page

Auteurs

Gianluca Gariani

Doctorant, C2RMF, Paris – Laboratoire de Physicochimie des Polymères et des Interfaces (LPPI), CY Cergy Paris Université – Fondation des Sciences du Patrimoine, LabEX PATRIMA, Cergy-Pontoise – Chimie ParisTech, PSL Research University, CNRS, Institut de Recherche Chimie Paris (IRCP), Paris (gariani.gianluca1[at]gmail.com)

Articles du même auteur

Charlotte Hochart

Ingénieur d’étude, département Recherche, C2RMF, Paris (charlotte.hochart[at]culture.gouv.fr).

Articles du même auteur

Nicolas Mélard

Conservateur, responsable des collections, musée d’Histoire naturelle, Lille (nmelard[at]mairie-lille.fr).

Articles du même auteur

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search