A big shout out to @EAAGlasgow (EAA Glasgow 2015). Would have loved to participated in this year’s timely sessions, but the Ryerson MDM grad students are starting early this week/year! I also want to congratulate my research partner Craig Barr who just recently launched his Lynda.com course entitled Mudbox Essentials Training. Mudbox is an extremely powerful 3D sculpting and painting software application, which we’ve also used in this project as well.
It has been a week of minor refinements and some time to review roofing structures for our longhouse. One of the key elements was to incorporate Bill Kennedy’s comments and observations in Longhouse 3.2. Bill has been building physical heritage reconstructions for years and had indicated that to keep the internal structure from bowing out over time, a lower mid-section horizontal support pole attached to either side of the bunking system would help keep the two sides from buckling out.Unlike the previous image below from Longhouse 3.2, with the addition of the mid-section horizontal cross-brace, the image of the Longhouse looked like it became very squat. Craig and I actually went through the settings in Maya to make sure the render from the camera wasn’t being pinched in any way. Artistically it clearly looks odd, however technically nothing has changed in height, width or length.
This “paradata” process of writing about the decisions made while producing a heritage object in virtual archaeology is a pillar of The London Charter discussed in Longhouse 3.1.5. It ensures transparency in the creation process,as the model assets are being built and how those assets are then applied in virtual space for public consumption. In my particular case, I’m using my artistic side of the brain to question if the visualization is correct, which should prompt an investigation on whether there is archaeological or practical knowledge data to back up my concerns. We concluded however that it was an optical illusion due to the change in wall post styles and the horizontal cross brace seemed lower to the ground level then what we would expect. For the next iteration it was decided to increase the height of the cross brace to roughly 60% of the total height of the longhouse. As discussed previously, we have been mixing building methodologies from Wright, Kapches and Snow along with insights from our commenters. The 60% height was suggested by Snow (see Longhouse 1.5) originally with the remaining 40% being a separate roofing structure completely independent from the wall posts. Wright suggested a similar approach but with a much taller wall system of almost 80% wall and 20% independent roof. Our hybrid model is going to incorporate Snow’s 60% wall height, then a continuous Kapches wall post and roofing methodology. The assumption being that native builders would have continuously tied down the exterior wall post framing to the interior support poles. Having the mid-section interior cross-beams and supports higher to about 60% of longhouse height, would also allow the exterior wall poles to be bent more naturally and one would assume it would then be more stable and secure?We’re also using Ron Williamson’s Fort Erie image below for inspiration with regards to exterior poles and what they would look like at the top of the roofing system. I’m going to assume that Iroquoian builders tied down the ends as opposed to cutting them. The image below has a separate flat roof supported by the interior support structure and scaffolded by the exterior wall system. This makes a lot of sense as the smoke holes were assumed to be square or rectangular in shape, which would be impossible to make if the ends of the wall posts met in the centre of the roof and were tied down. Ultimately we have no clue whatsoever how the roof was made, so we’ve decided to make a flat roof supported by the interior rafter system then round off the roof at the top with the 1x2m shingles that will form the outer shell of the house.
The image below is a revision on the notes discussed above. In it we increased the height of the cross-beams and then also changed the exterior curve of the wall posts. I’m envisioning two major tie down points for the exterior wall at the mid-section and top of the rafters. With all of the potential tension on the mid and top sections of the wall posts, I’m also wondering if they tied down the bottom of the wall posts to the interior bunking system at the base? Generally the poles would have been placed into the ground at about 1 to 1.5 feet. Sometimes individual holes were dug, but there is also archaeological evidence that slip trenches were dug as well and then dirt filled in afterwards. In the previous rendered image above, Craig randomized the placement of the posts, similar to how we find them in the archaeological record which suggests a roughly straight line, but not precisely straight. Lastly, following Wright’s observations on pole height vs taper, we applied a greater taper and random length to the poles.
As with the normal model building process in 3D, we tend to forget or leave important corrections out while revision notes come through. In Longhouse 3.2 I noticed a lot of texture and modelling issues that were just simply items to be cleaned up. Floating support beams, some bunks without supports and areas with no rope strapping.
By chance we got our Ocular Rift Dev Kit 2 this week. Craig quickly imported the 3D layout from Longhouse 3.2 in Unity 5 and then donned the OR to do a virtual “site inspection”. By walking around the structure in 3D, he was able to quickly pinpoint all of the areas that needed to be remodelled or cleaned up. It was definitely a unique experience to check for issues.
One of the elements you may not have noticed in the image above is the creosote texture mapping on the upper rafters/support beams. Unfortunately a lot of this detail will get lost once the longhouse is enclosed and the lighting added, but we felt it was a necessary detail to add in terms of discussion points later in the process.Lastly Craig hand adjusted every rope lashing point in the longhouse. Typically in 3D animation, we make one model of an element and then copy and paste that 3D model if it’s going to be repeated. In this case, Craig remodelled every rope element as even the pole and post diameters were randomized to present a less uniform 3D look.
Our next iteration will have the exterior shingles in place and the cedar flat faced wall that was used to separate the rounded storage/entrance vestibule with the house. As always, any comments, questions or insights are greatly welcomed.