photogrammetry – Early Medieval Making

ORP: Eriswell Spearhead

June 4, 2018 pearsonh Leave a comment

Little Eriswell

In 1957, workers digging a hole for the installation of an oil tank discovered bones and artifacts indicative of an Anglo-Saxon cemetery at Lakefield airfield near Little Eriswell in Suffolk. Two years later, another patch of the cemetery was discovered while foundations were being dug for a hospital. An inventory of all the findings, including objects that were buried with the people there, was kept during the process of the excavation. In one of these graves, an iron spearhead was recovered, found well above where the left shoulder of the body would have been and with wood remains still inside the conical ferrule, the section which would have affixed the spearhead to the shaft of the weapon.

The original spearhead from the Eriswell excavation.

Spears in Anglo-Saxon Culture

Spears were believed to be incredibly prolific weapons during the time of the Anglo-Saxons, due to the majority of them having been found within burial sites; roughly 4 in 5 burial sites contained one. They were much cheaper than swords to produce, and so they were frequently allotted to foot soldiers. Anglo-Saxon spears themselves definitely had certain distinctive qualities to them, such as open ferrule sockets around the spear’s mandrel, which would lead to such an assumption. It is possible that this was the role of the particular individual who died, as also found with the remains were a shield boss, a pattern welded iron sword and a small iron knife, though it is wise to remind myself that the dead do not bury themselves. In any case, it draws a strong connection to spears being symbolic of a warrior. Despite this, archaeological data on weapons burials does not corroborate this notion.
Weapon burials were, according to research, heavily associated with the “ethnically, social and perhaps ideologically based ‘warrior class’”. This was seen as many burials that included weapons were also frequently adorned with generally far more wealth than those of adults that had not. Skeletal data also suggests against the idea of the simple fact of death in battle or of being a warrior would qualify one for such a burial; there was little correlation for wounds and weapon burials and some weapon burials even had people with various diseases and disfigurements that would have precluded their service in an army. The evidence at Little Eriswell absolutely corroborates this; among the spearhead was also found several more weapons, including those of a very advanced make, including a pattern welded sword, a quite expensive and masterful piece of equipment for the time. Furthermore, the skeletal parts that did remain showed several vertebrae with Schmorl’s nodes, a condition that denotes spinal deformity, as well as evidence of osteochrodritis dissecans on the right femur in the form of a shallow concavity, a disease that causes stiffness and locking of the joints: certainly not a boon on the field of combat. While there was not much commentary on the metalwork of the spearhead itself in the Little Eriswell report, there is some implication that it was a least somewhat well made, with iron fragments of other items having been found in the same grave, perhaps implying that the spearhead was at least made durable enough to last the test of time.

Making, Part I: Photogrammetry

I had no experience in metalwork, nor the time to learn such an art. So I “made” in the way I knew how: digital modelling. My first attempt was through a process called photogrammetry, where I was given photos of the artifact itself on a green screen from many different angles, where I then tried to construct a digital model using a computer program. The first step in this process was called masking, where I took the 37 photos of the spearhead and cut out parts of the background that I didn’t want the computer to use in the final construction. I then began to run the program through the process of building the model from these photos; aligning the photos, building a “density cloud”, which creates a cloud of points that the program will later fill in when it creates the mesh, or the digital 3D object, and then finally building the textures, which includes all the colors and finer details that are too small to perceive properly, so they are set as variations in color in order to make a less resource intensive final product. Unfortunately, after repeating this process over and over, troubleshooting many errors and watching tutorials, as well as seeking help from other students, I was left with nothing, due to the quality of the original photos that had been taken.

The program, Agisoft Professional, and the beginning vertices of the spearhead.

The density cloud. Note the lack of detail in the spearhead itself, and mostly in the mat beneath it.

The final model. Note the correspondence to the density cloud in terms of detail.

The most frustrating part about this failed attempt at recreation was how hands off it had been. Unlike actual metalsmithing, this approach was far less intuitive and automated; in fact, it would be less accurate to say that I had made the model rather than communicated to a computer what I wanted, which then had made the model. This led me to seek a far more hands on and involved approach, while still utilizing skills that I possessed.

Making, Part II: Freehand 3D modelling

The next process that I pursued was that of freehand modelling. This process consists of using a digital workshop within which a 3D artist can align points by hand. A 3D model is primarily built from flat triangles, which themselves are composed from 3 vertices, or single points in 3D space, connected by 3 edges, which are in turn connected by a single “face”. All this data is conveyed numerically to a renderer, which projects little lines in 3D space; if it would hit a face, it renders a solid surface there. These rendering lines are drawn for every pixel on the screen; this is why higher resolution screens generally take up more resources.

What I did was I used the various tools provided to me by my program of choice to create what I thought, based on research, this spearhead could have looked like when it was first made. So I first decided to create the head, by first creating a rectangle, then deleting and reshaping the geometry to create the leaf-shaped figure. I then “extruded” or lengthened the back of the head, and created a conical socket by folding the vertices into themselves. I then separately created an admittedly lazy “shaft”, which consisted of little more than a stock cylinder that had been stretched out like a piece of taffy. I placed this shaft within the cone, and using a “sculpt” tool, I “hammered” the edges of the cone around the shaft, making sure to leave the joint open in true Anglo-Saxon style.

The spearhead, before use of the sculpt tool to close up the ferrule

This is where the simulation of the method began to fall apart. First off, the model was far too flat and blocky, even though it held the general shape quite well. This meant I had to apply what is known as a subdivision surface, which essentially adds detail and a smoother overall shape to the model. Furthermore, both pieces were colored a very dull gray, which is standard for almost all modeling programs, but I wanted something with color. And so I mapped the faces of the model on some stock images of metal and wood for the head and shaft, respectively, which allowed them to display the respective portions of the image within the renderer. I tried my best to mimic how the metal and the wood itself bent around the shape of the spear, but without much hands-on experience with the relevant craftwork, I found this quite difficult to ascertain. There were also lighting issues being caused by some of the unseen geometry, such as within the socket, and so I had to delete it, leaving the inside of the mesh completely transparent (only one side of a face actually renders solidly; in order to have both an inside and outside to an object, two faces must be created). This forced me to close up a bit more of the joint, and flatten the spearhead even more to the shaft. Finally, I was also having issues with the aesthetic qualities of the spearhead; that is, it was far too reflective, not only looking more like a polished plastic than steel, but also washing out a lot of detail, such as the spine of the blade.

I then took this final model, and with the help of Brittany Johnson and Austin Mason it was 3D printed, and this was the most satisfying part of the experience, because I finally got to touch and feel the object that I had been creating. It wasn’t perfect; it was only about the length of my hand. But it still gave a tactile sense of this object, which is important to understanding the lives of those who used it.

The 3D printed spearhead. A little smaller than the real deal.

Conclusions from the Experience

The advantages of 3D modelling were obvious enough from the get-go. The infinite material, as well as the tools that allowed for an almost limitless range of geometry and incredibly efficient workflow meant that I could produce the most functional part of this object very quickly; its shape. One thing that had occurred to me, especially during troubleshooting and tweaking, is that the intent of my design was quite different. Indeed, I was focused more on the aesthetics i.e. the shape of the spear head, the geometric qualities of the ferrule, the reflective qualities of the metal, whereas an Anglo-Saxon blacksmith would have been more focused on function. If it were easier to not have the joint be separated when smithing a spearhead, it is unlikely that we would see the split ferrule on these spearheads.

The final product.

This, I feel, highlights the value to this sort of modeling: when I tried replicate the object, as well as ape the methods used to create, I was forced to put myself in the mindset of the smith. While maybe I didn’t learn a similar appreciation for technique and limited material that the actual craft would have, considering shape and look so closely made me feel a little closer to the material culture itself, and as a result made me feel closer to the actual people, allowing a more accurate interpretation of the culture based on the things they used. The specific value of 3D modeling is its ease and accessibility in comparison to many of the crafts that would have been practiced at the time, in particular metalsmithing.

Tag: photogrammetry