In my last post, I discussed the DiGA project with two of its members – Dr Jessie Pons, project leader, and Dr Cristiano Moscatelli, research associate. We chatted about the project’s incipit, logistics and management. In this post, our talk focuses on the practicalities of digitization. We are also joined by Dr Frederik Elwert, DiGA project leader.
Alice Casalini (AC): What kind of equipment did you use for the photos? Were there any challenges about working with sculptures/3D objects?
Cristiano Moscatelli (CM): We couldn’t set up a proper photographic studio in the museum due to logistical reasons. As a compromise, we used two lightboxes, which come with some limitations. For example, positioning objects, especially fragmentary or large pieces, can be challenging, and the size of some objects often poses a challenge. This is why we got two boxes of different sizes, one suitable for pieces with a max. height of 60 cm, and the other for objects with a max. height of 80 cm. At any rate, the lightboxes proved to be a good choice. They came with dimmable light bars that were easy to adjust, so light control wasn’t a major issue.
Besides lightboxes and cameras, our equipment included tripods, scales, black velvet backgrounds, labels for noting down objects’ inventory numbers, color charts for adjusting white balance in post-production, and acrylic props to support the objects. I used a 50 mm lens which, compared to other types of lenses, produces minimal distortion — which can be easily corrected in post-production.
Tools of the trade. ©KPDOAM/DiGA CERES.
Some objects, however, were so heavy and large that placing them inside the box was a real challenge—and sometimes even impossible. I remember two pieces in particular: one was a Buddha statue, chipped in various spots, which could not be put on a base. I ended up using a couple bricks we found outside the museum, wrapped in black fabric, to create a support at the back (necessity is the mother of invention…!); the other was a pair of umbrella disks, which were quite heavy and didn’t fit in the lightboxes. We had to change the entire setup. We placed the objects on the floor, and museum workers helped hold the lights while I set the tripod on a table, with the camera pointing downwards.
©KPDOAM/DiGA CERES.
As for the setup, three specifications come to my mind that some might regard as trivial but are actually quite important. First, I would ensure that I am shooting in RAW, which is the only format that retains all the image data and can be processed later without losing quality or information. Most importantly, RAW files serve as proof of authenticity. As they provide an unaltered original, much like an old negative, they prevent intentional manipulation of the content. Second, I would check that ISO values (sensor sensitivity to light) are set to the lowest possible, to avoid images full of digital noise. Third, the use of a delay timer is fundamental. Even if the camera is mounted on a tripod, pressing the shutter button can result in micro-motion blur. A delay timer is a good way to avoid this.
As for lighting, it really depends on the type of object. If lights are not properly set, high-relief objects might cast heavy shadows, while low-relief objects might appear flat. I would mostly use 45° lighting, and occasionally add lights on either the top or bottom of the box or even in front of the object, adjusting their position and intensity depending on the result I want to achieve. It’s a subtle dance…
Stucco heads. ©KPDOAM/DiGA CERES.
AC: Did you do any 3D documentation?
Jessie Pons (JP): Yes! Originally, we had planned to use photogrammetry for the 3D capture. Instead, Calin Suteu, our service provider, used a 3D scanner (Artec Spider). This significantly accelerated the scanning of the objects. Objects were scanned for 3D models for the first time during the 2023 field trip. A 3D station was set up in one of the galleries of the museum, which consisted of a table with a spinning support to hold and maneuver the artifacts during scanning. Artefacts selected by the DiGA team prior to the field trip were gradually brought to Calin Suteu by the museum personnel and laid on the spinning base. The scanning was performed in several stages, covering all the faces of the object and the post-processing was partially performed in situ. Calin Suteu managed to scan 71 objects. The models were then exported in commonly available formats (.OBJ, .STL). A sample can be seen on Calin’s Sketchfab account. Others have been uploaded on heidICON. You can read more about our 3D process here.
3D Documentation. ©KPDOAM/DiGA CERES.
AC: Cristiano mentioned some of the issues regarding the size of some of these sculptures. Did you encounter any other unexpected difficulties in working with these objects?
JP: Calin mentioned that the type of stone (mica schist) made his work a bit difficult. The mica flakes tend to reflect light, which interferes with the laser. Sorting the objects by type and size beforehand would have helped with the setup of the photographic framing for several batches of objects in a row, rather than having to change settings often. Another issue I’d like to mention is related to restoration and conservation. Some of the objects had been restored in the past after coming to the museum, but materials that are inappropriate for a modern standard were used, like cement or glue and tape; others were badly chipped and very encrusted. We did the best we could with the resources at our disposal, but some objects should have been cleaned and repaired by a professional conservator before digital documentation. This is something to keep in mind for future projects: the digitization phase should always be preceded by a thorough assessment of the collection with these criteria of conservation in mind.
AC: Who managed the storage of all the data and images? How were the digital images managed in terms of moving them across different media and platforms?
JP: The data is stored on several platforms. The main platform is heidICON, which is our official repository. But the data has traveled quite a bit before it was uploaded to this database. After each photographic session, the data was transferred from the SD card to an external drive. The data was then checked (are the images sharp enough, have all sides been photographed, etc.). The photos were then post-processed. Aurangzaib Khan [who joined the project as a professional photographer] and Cristiano edited the photos they took during the first field trip, but for the remote campaign and the second field trip, we Claudia Primangeli, a third-party service provider, to speed up the process. We mostly used external harddrives to send the photographic material. The data was too large to use data-sharing platforms and uploading/downloading material took too much time.
Frederik Elwert (FE): In many cases, physically mailing well-packaged hard disks prove to be much quicker and more reliable than any transfer over wire! Managing the files proved to be a bit of a challenge. We had a quite detailed plan for the folder structure, file names, etc., but the actual situation in situ made it difficult to always follow them to the letter. For example, sometimes we realized that one side of an object had not been photographed at all, or there were some quality issues. We re-took the photo the other day then, but this sometimes led to duplicates that had to be sorted out later. Some data was also spread across multiple harddrives, as people were working on them in parallel. This sounds trivial, but we spent quite some time checking file names and resolving duplicate entries.
JP: The photographic material is also stored on Sciebo, the Ruhr University Bochum’s cloud service, and will be archived on heiARCHIVE. A set of the photographic material was also given to the Directorate of Archaeology and Museums of Khyber Pakhtunkhwa (KPDOAM). As for the metadata, and more specifically the descriptions, we created word files and shared them between Cristiano, Serena [Autiero], and me for correction.
AC: Can you tell me something about the process of creating this digital archive hosted on heidICON?
FE: Using this kind of pre-existing infrastructure has its pros and cons. Our data are now somewhat hidden in a large collection of images from different contexts. Also, the presentation is quite technical and does not invite simple browsing of the collection. Some features we would have liked, like the annotation of iconographic features on the images themselves, are not yet available on this platform. However, one of our main concerns was that the chosen infrastructure would still be operational even after the end of the project. Our data is available at heidICON for the foreseeable future – this is a big benefit for institutions like university libraries. Starting from there, we can still experiment with different modes of presentation and re-use parts of the collection in different formats.
AC: My understanding is that DiGA has concluded. What is the future of the project, if any? Are there any other similar projects you are involved in?
JP: The project has officially concluded, but there is still work to be done on the database. We’ve identified some duplicates, discovered errors in a few descriptions, and we still need to upload and update some of them. Before we can consider the project truly “complete,” we must undertake a thorough round of proofreading and “quality control.” One significant milestone we are trying to achieve is to ensure that KPDOAM has its own local version of the database: right now, they have the photographic material but lack a presentation platform. We are currently exploring solutions for this.
We have other ideas for potential follow-up projects. These include a systematic study of the corpus to reconstruct the history of Buddhism in the region; the development of public outreach strategies and capacity-building programs; the application of our digital concept to other collections under KPDOAM’s jurisdiction. It has been an honor and a pleasure to collaborate with the Italian Archaeological Mission (MAIP/ISMEO), KPDOAM, and Heidelberg University Library. We would like to extend our collaboration beyond the scope of the DiGA project. As always, the feasibility of such collaborative research depends on funding opportunities, but where there’s a will, there’s a way.
The Digital Orientalist 