The following are the main points of my PhD proposal. You can also check out my profile at the Glasgow School of Art for more information.
There are two major issues surrounding small finds in professional archaeology: 1) access to the objects and 2) the creation of visual records appropriate for remote archaeological analysis. Recording approaches are complicated by the reflective and/or translucent materials used to create many small finds. I believe 3D modelling has many benefits for visualising objects that prove difficult for traditional methods. Yet, to date little has been done to create models of small finds appropriate for use by specialist researchers.
This research will investigate the uses of 3D modelling for small finds, particularly beads made of glass, stone, amber, and faience. There are two primary objectives: 1) Find cheap, practical, and easily replicable methodologies for creating 3D models of small, translucent, and reflective objects and a database within which they may be compiled.; and 2) Make the models and related data available in an open-access, relational database. This will culminate in a detailed methodology useable by other researchers. The ultimate goal is to bring new technology and techniques into small finds research and archaeology as a whole.
This research is important for several reasons. First, 3D modelling will create affordable, practical methods for visualising and accessing objects that are otherwise challenging in those regards or simply unavailable. Second, the methods developed to solve issues of translucency and reflectivity can be applied to any object with these characteristics, archaeologically or otherwise.
Beads are one of the most common small finds recovered from archaeological sites worldwide, with certain sites recovering thousands each excavation season. They have been one of the most ubiquitous trade items for at least 2500 years. Yet, there are few if any standard practices for recording and analysis, making research difficult.
No systematic method for classifying beads existed prior to the 1920s, when Beck (1926) published his typology categorizing beads based on shape alone (1926). Archaeologists generally used Beck’s system until the early 1970s, because it was an objective, measurable classification system with few, if any, alternatives. Yet, in 1973, van der Sleen (1973, 51) insisted that other physical qualities of the beads (like colour) and the socio-cultural context of the beads were equally important. Various typologies also began emerging in specific regions around the world. Guido published her seminal work on Iron Age and Roman glass beads in Britain (1978) and Callmer published his on glass beads in Scandinavia from 800 – 1000 AD (1977). Francis also began publishing his numerous catalogues of Asian material beginning in the early 1980s (1982a, 1982b, 1988-89, 1990).
These typologies are useful within the regions in which they were created, but they are often difficult to access and do not allow for meaningful cross-cultural comparison. Furthermore, site reports often only mention how many beads they have, or possibly illustrate one side of one example in a scaled drawing. They rarely discuss bead types in any detail, if at all. This and the issues of illustrating such small objects in photographs or scale drawings lead to difficulties of accessing any information about beads without traveling to the museum in which they are housed to view the objects oneself. Such a venture is often costly and time consuming, precipitating the need for digital access to 3D models.
3D modelling is a rapidly maturing technique in archaeology. Photogrammetry as a process has been used in archaeology primarily on large statues, buildings, and even cities (Bernardini et al., 2002; Grün et al. 2004; Remondino et al., 2009; Remondino 2011). However, photogrammetric software struggles to process images of translucent or reflective objects, which complicates its use for objects like beads.
Techniques have been developed for processing images of reflective and translucent objects to reduce these effects, such as using polarizing lenses and multispectral imaging. In archaeology and museum settings, these techniques have been used primarily on documents and paintings (Cain and Magen, 2011; Chabries et al., 2003; Liang, 2012; Pelagotti et al., 2008). Recently, however, some researchers have experimented with using image pre-processing to create files that photogrammetric software can recognize and compile into 3D models (Guidi et al., 2014; Nicolae et al., 2014). This has worked relatively well with objects such as marble statues, ceramics, and metal vessels. Yet, to date there have been no studies applying these methods to smaller objects, particularly small, translucent, and reflective objects.
For my MLitt research, I documented and analysed over 1200 beads from Scottish contexts dating to the first millennium AD. Since I am already familiar with this material, I will again focus specifically on beads from Scottish contexts dating to the first millennium AD.
The primary goal of this PhD is to examine existing methodologies for visualising small, reflective, translucent objects (specifically beads) and to develop new methodologies where needed. It will therefore concentrate on the quality rather than the quantity of the models produced, and the process through which those methods developed. I will apply these methods to as many beads in as many collections as possible in order to test various materials, styles, shapes, and other complexities often found in beads. These include multiple layers of material, inclusions, bubbles, and breakage or corrosion.
I will create the 3D models using a number of techniques, the primary one being photogrammetry. Since most materials used to make beads are both translucent and reflective, I will process the images prior to creating the models. Adjusting luminance settings through photoimaging software can significantly decrease reflectivity. HDR photography and polarized lenses can also significantly reduce or eliminate reflectivity. I will also use microscopic imaging with polarization to achieve as much detail as possible on such small objects.
I will then test the uses of multispectral imaging and reflectance transformation imaging (RTI) in creating 3D models of small finds. These techniques measure the reflectance and fluorescence of pigments. The photographs taken using each of the spectra provide different information, much of which is invisible to the naked eye under normal lighting. 3D models created with multispectral imaging and RTI could therefore provide much more information than that obtained by more typical methods. I will also experiment with creating 3D models of subsurface features, such as bubbles, inclusions, and separate layers of glass.
The process for creating these models will culminate in the development of a workflow usable by other researchers. This workflow will explain the developed methods in detail to allow for widespread use. It will also detail any issues or problems encountered within the development and use of this methodology and how those issues were solved.
I plan to compile all models connected to each bead in a single location, using filters of multispectral and polarized imaging to be turned on or off as layers. I will then connect these models to all other contextual information regarding the bead through a relational database. I plan to make this database as widely available as possible online, given the proper permissions.
Ultimately, the goal for this research is to create portable, affordable methods for visualising small, reflective, and translucent objects (or objects with any combination of those characteristics). While the immediate use of this technology will be in the field of archaeology, it can apply to any object with one or more of these characteristics, and can therefore transfer to any field requiring such visualisation techniques.