Discussion

Overall, though we ran into some bumps along the way, we do feel that this project was a success. Our initial intention was to increase the ways in which people could interact with the Perlman's collection, as well as to provide the museum with more information about the artifacts themselves; both of these goals were achieved, in that we produced 3D models of the majority of the Perlman's Greek collection, and clear photographs of the rest, as well as research into the origins and uses of these artifacts.
Furthermore, we were able to explore various tools for 3D modeling, as well as website creation and artifact research, all skills which we hoped to expand during the course of this project. This was possible despite the various issues that arose during the modeling process, which forced us to pivot from using MetaShape to Scaniverse. More specifically, we began the project by photographing the artifacts and using those photos in MetaShape to produce 3D models. However, we encountered problems with aligning the photographs, not having enough photographs, the lack of space taken up by the object within the frame, and issues with the coverage provided by the photographs we did have, all of which resulted in unusable models of several of the artifacts. As a result, we decided to use only Scaniverse to model the artifacts, to ensure that we would be able to check the models while we still had access to the artifacts (as Scaniverse is a mobile application, we were able to generate the model within a few minutes of creating the scan, meaning we did not have to wait until the artifacts were re-stored and the models were generated using the MetaShape software at a later point).

While the use of Scaniverse did produce more usable models, it was still not without its own problems. Most pressing was the fact that some of the objects we were attempting to model were quite small, and had to have various levels of support. Though Scaniverse works very well in many ways, it is not the best platform for modeling very small objects, and we realized that some objects simply would not produce workable models. As a result, four of the objects have only photographs, rather than 3D models and photographs. This was a disappointment, but we were still able to photograph and research these artifacts, allowing for some new resources to be provided to the Perlman.

We also noticed some patterns in the modeling process, in terms of which objects modeled well and which produced more difficulties. In general, the bigger the object, the more likely it was that a usable model was produced. This was especially true when using Scaniverse, which was totally unable to register artifacts as small as the copper ladle, for instance. Additionally, matte objects were more successful than shiny objects, as they did not bounce light off their surface in a way that made it difficult for the modeling program to determine the edges of the object. Metal objects also produced slightly messier models than the terra cotta objects, as there tended to be more deterioration of the material which produced much more texture variations for Scaniverse to deal with. On the whole, we found that non-shiny terra cotta objects produced the best models, while metal objects still produced clear and workable models, and that size was an important factor in the success of the modeling process.

Lastly, its important to mention that creating the website was relatively easy for us as we had two CS majors in our group who had experience with creating websites and knew what technologies could be good for this. That is why we created our Figma prototype since one of them thought it could be a good idea to make one before we started writing information for the website. Additionally its also how we knew to make the website using Bootstrap, whilst it was a bit tedious to make sure that everything was formated properly (such as hyperlinks that don't automatically have the necessary tags). It was still a welcome challenge to partake in.