Models of the Triple D Kind

This week we tackled the big field of 3D modeling. 3D models are becoming more and more prevalent in today’s computerized society. 3D models can tackle many different varieties of research questions, ranging from preserving an archeological site to city planning. 3D modeling is most useful when it is applied to projects that require the use of 3D space in order to fully visualize the solution. You can use 3D modeling to tackle most problems that modeling already tries to tackle with even more effectiveness.

The field of modeling is already a well flushed out field, 3D modeling is just a another addition, albeit a large addition, to the field of modeling.

An interesting note on 3D modeling is that there are different processes of modeling in the 3D space, and there are different ways of generating these models. 3D modeling includes, procedural modeling, photogrammetry, and scanning. I found procedural modeling most interesting because it can create such complex bodies and models only using a simple foundation of rules that tells the program how to generate the model. By using a simple set of rules programs can generate such complex cities and structures without much help from the programmer. Procedural modeling is most effective for large models, for example for city planning procedural modeling makes the most sense because you don’t have to manually model every single tree and sidewalk in a large city. But procedural modeling falls short when it deals with more intricate and specific types of models, it also falls short in the hardware department because it takes a lot of computing power and time to procedurally model anything. Scanning and photogrammetry makes the most sense when you want a very detailed model of an object. For example the preserving or digitizing of old relics in a museum would be great to utilize scanning or photogrammetry, because these modeling processes will capture all the small details that one needs in order to fully preserve a relic. However it falls short on anything much larger, because larger things have more vector points to scan and by increasing the amount of vector points that you need to scan it would take a lot longer and a lot more processing power to chug through all of those data points.

This week we explored Marie Saldana’s DH project as an example of how 3D modeling can be used in projects. Marie used procedural modeling to generate Roman cities. Since Roman cities very much so followed very structured rules on how they layout their cities it was her goal to write out the set of rules to procedurally generate Roman cities. It worked very well in that it gave very nice rough estimates of what a Roman city would look like without actually manually modeling all the intricate details in such a large city. However a shortfall in procedurally modeling is that since the city is created using a set of rules the program follows, how do we avoid the “trap of generalization”. As Marie puts it

“Another challenge is rule-based modeling’s inherent bias for finding isomorphism and ignoring singularities. In other words, how do we deal with the ubiquitous cases where the so-called ‘rules’ turn out to be broken?”.

Procedural modeling is also very limited in its choice of software. For Marie’s project she only used CityEngine for it and no other software, this just shows the scarceness of software that can procedurally generate models.  CityEngine also has its short comings, as Marie found “[CityEngine is] currently being targeted and developed for the urban planning, production design, and gaming markets, and is therefore not necessarily optimized for scholarship or teaching.”.

Overall I thoroughly enjoyed this week’s material. 3D modeling is a huge field that will be invested in a lot in the future due to its versatility in solving research questions. Not to mention it’s also really fun to mess around with 3D objects on the computer.