You need a circle to build a house

The plan of the House of Orpheus was the first one I drew directly in 3D studio max. 

Initially, I considered using Adobe Illustrator (a software I used quite often in my previous job) but then I realised that now I am more comfortable with Max than with any other graphic software. Furthermore, a plan drawn in Max will be entirely compatible with the 3D model (avoiding some of the problems that imported files may cause).

My plan of the House of Orpheus in 3D Studio Max

At the beginning, I was quite clueless about the methodology to use but Drew, my supervisor, thought me a very effective method based on an initial assumption and the use of circle shapes (or 25 sided n-gon which, in this context, is pretty much the same).

Our initial assumption (arbitrary as every assumption) was that the atrium of the HoO is a regular rectangle. Although I really doubt that there is any really straight wall in an ancient Roman house, the assumption is mandatory as a starting point. Moreover, the atrium is actually a fairly regular space.

Then, using the data we collected on site, I designed circles with the same ray as the dimension I wanted to represent. The very handy thing is that, being every point of the circle at the same distance form the centre, the circle is still a precise mark even when the adjacent element turns out to be not completely aligned.

Why did I need such a feature? They say numbers never lie, but in my experience they never completely match either :-)

Material errors during transcription, different layers of plaster, non regular shapes. I can think of a number of variables that may affect the measuring operations. Or it’s just a kind of useful curse to remind 3D modellers that they are not reproducing “the real thing” but only a representation of it. Actually, I encountered a similar issue when I was working on the Temple of Isis. Not only I have never seen two plans with identical measures but sometimes it was difficult even to find two sources that roughly agreed on something. Now I can easily imagine the headaches that Francesco Piranesi or John Soane suffered!

My plan of the House of Orpheus in 3D Studio Max
(all the working layers visualised)

Considering the basic level of fuzziness (usually between 2 and 7 cm) embedded in my data, using circle shapes was extremely useful and allowed me to deal with that fascinating jigsaw of walls, doors and columns.

Here are few things I should say about this digital plan, in order to make it more transparent.

* The thickness of the walls in the House of Orpheus if fairly regular. With the exception of the wall separating rooms 5 and 10, they all measure between 40 and 45 cm. Thus, I thought it was safe to assume that the walls I wasn’t able to measure because they are shared with other houses we were not allowed to enter (such as the south wall of the garden, or the west walls of rooms 14 and 15 ) follow the same rule. I also made the assumption that the walls overlooking the street are a bit thicker than the internal ones.

* The length of the viridarium (from south to north) has been calculated summing the diameters of the columns to the distance between them. Even though I have measured every single intercolumnation, I have measured only one column and I have assumed (for simplicity) that they have all the same diameter. I also assumed (again, for simplicity) that all the columns were perfectly aligned. My calculation was quite well backed up by the sum of the elements belonging to rooms 12, 4, 5, 10 and 11.

* All the traditional plans of Pompeii I have seen so far, tend to simplify the shape of the rooms and make them more regular than they actually are. I have chosen to preserve the irregularities and the oddities in the rooms dimensions because I thought they might be relevant in the visual analysis.

* The positioning of the impluvium in the present map is an approximation. It was the last element of the House we measured and we were running out of time (and light). I measured the dimensions of the impluvium itself and its marble elements but I didn’t measure their distance from the walls of the atrium.

My supervisor is quite confident that we could derive a more precise position through the photogrammetry of the atrium he ran.

* When I had little inconsistencies with my data I chose the option tending to the most regular shape.

Orpheus has a new plan

VanderPoel's plan of the House of Orpheus
Highlight in orange: the south wall of the garden

In order to model a visualisation of the House of Orpheus, I decided to make a new plan of the building, based on the hard measurements we took on site.

Why? What is the added value of such an effort? Wouldn’t have been more sensible to start from a good pre existing plan, such as Presuhn’s or Vander Poel’s ones?

I had basically two reasons for that.

First, maps and plans are simplified visualisations of places. Thus, they are «models», ie artificial objects created to represent and study reality. This implies that they cannot represent all the aspects of reality but have to choose what is relevant for a specific purpose. To make it simpler, I wanted a plan made ad hoc for my purpose, and the only way to have it was, basically, to make my own.

For example, if we compare Presuhn’s and Vander Poel’s plans we’ll notice that they are slightly different. In the latter, the south wall of the garden is slightly divergent from the north one, while in the former the south wall converges to the north one quite dramatically.

Who is right and how did it happen? In my opinion, the point is that all maps start from some artificial assumptions. The first assumptions affect the entire visualisation of the data. Presuhn and Vander Poel probably started from different assumptions, thus their plans (which are both not «the real place» but a representation of it) are not identical. However, if their purpose was (as I think) to visualise the relationships between buildings and their arrangement in the city plan, the maps are both useful and both «correct».

Presuhn's plan of the House of Orpheus
Highlight in orange: the south wall of the garden

However, my purpose wasn’t to examine the House of Orpheus in a relational perspective. Actually, it is very relevant to me if a wall like the viridarium's one is straight or not because it would definitely affect the information about visibility of spaces.

Eventually, I decided to build my own map relying on my hard measurements, photographic documentation, previous plans and satellite photographs as the ones offered by google maps and google earth.

Second, Johanson says that there is no better way to understand an ancient building than rebuild it, at least in a virtual space. I discovered it when I started modelling and this relationship still deeply fascinates me. In my previous experience, I always started from someone else’s plan. At the beginning, I saw the need to draw my own one “only” as a challenge and an opportunity to learn new skills. 

Then, when I started working on the plan I realised how much I was already learning about the building itself, the relationship between spaces, the orientation, the proportions, the position of doors, the balance (or lack of it). When I finished the plan, I already knew crucial information about the building even before starting the proper modelling.