2018 Revival, 5.12-11.2. Week Recap, 12.2-18.2. Plans
This past week went mainly in spirit of Gothic architecture, and was capped off by sculptural development of the main character and start on a rig. Going through each topic, we will take a look at the workflows used, problems encountered, and solutions.
Gothic Architecture
During the first half of the week, I spent time on researching and developing (R&D) the methods of properly and truthfully capturing and constructing the Gothic Cathedral. A list of inspirations for the interior of the cathedral was compiled, and from it a single cathedral interior was selected as a main reference. The one selected was Duomo di Milano aka Milan Cathedral.
![]() |
Duomo Di Milano Interior Image by anpalacios on flick |
Research and Iteration
One of the first things to figure out was how will be the Vaulting constructed. Few alterations to the cathedral layout was made, mostly rounding up numbers from very awkward measurements and adding pillars where needed.
As can be easily observed from the picture, an essential part of vaulting and Gothic style is the prevalence of pointed arches. Now, there are 4 types of pointed arches in Gothic architecture: Lancet, Equilateral, Depressed(or sometimes referred to as Tudor) and Flamboyant. For our needs, we focused on the first two types, that being Lancet and Equilateral.
Equilateral Arch
As it's name suggests, equilateral arch is based on equilateral triangle, where the radius of the arch equals the side of a triangle between the point and base. It is drawn by constructing two ellipses on each side of the base with their distance equalling the radius, with the point of their intersection being the highest point of the arch.
Lancet Arch
Lancet arches are typically defined as two-centred arches whose radii are larger than the arch's span.
Equilateral and Lancet arch springing techniques Image by SBE Builders |
Turn up the Volume
One thing, that was of concern was loss of volume. To put it in the context, the workflow for plenty of hard surface meshes I choose is a SUB-D at render time. Therefore, polygons get subdivided by multiples of 4. For construction of the arches, I used curves, therefore for texturing, sculpting and rendering I needed a way to convert them to polygonal objects. Maya offers a variety of options to tweak and tinker when converting NURBS surfaces to polygons, therefore I tested some of them out in comparison with the polygon count that resulted from conversion, aiming for the best Value.
![]() |
Top: Pre-subdivison Bottom: post=subdivision |
Judging from the results (given they are atleast somewhat visible), using the Control Points option gives us the best results, both in maintaining volume, point being decently pointy and a very easy to handle amount of polygons.
Construction and Vaults
We constructed arches in a similar manner inside Maya, according to our measurements. Having completed the first piece of the vaulting, my research took me towards figuring out the proper shape and construction of the Gothic vaulting.
![]() |
Arches and Vaults measurements for the cathedral |
Furthermore, there are different vaulting styles, I mainly focused on these there:
![]() |
Different Vaults, Source Various |
Here is where I encountered one of the first major problems: The way my arcs were constructed, and the way I wanted them prevented the usual way of just running two different arches through each other and constructing the vault using Boolean operations.
Therefore, I had to find a way of construct diagonal ribs and another arch in the webbing from the smaller vault towards the intersection of the diagonal ribs. Normally, it would not be a problem, but according to my calculations which I checked multiple times, the orientation of the ellipses that would make up the diagonal ribs were very long numbers with a multitude of diagonals. Therefore I had to figure out a way how to snap them properly, which I in the end accomplished.
Therefore, I had to find a way of construct diagonal ribs and another arch in the webbing from the smaller vault towards the intersection of the diagonal ribs. Normally, it would not be a problem, but according to my calculations which I checked multiple times, the orientation of the ellipses that would make up the diagonal ribs were very long numbers with a multitude of diagonals. Therefore I had to figure out a way how to snap them properly, which I in the end accomplished.
Another smaller problem, was that previously used equilateral arches did not fit, therefore the diagonal and webbing arch are constructed using lancet arches, which was another interesting endeavour.
The very last problem I had encountered was actual creation of the webbing. By my workflow, I had a variety of NURBs curves, since the plan was to use them to create an accurate surfaces. One half of the surface ended up being accurate, but the other half left me wanting for more control, therefore in the end I constructed the webbing using traditional polygonal techniques.
The very last problem I had encountered was actual creation of the webbing. By my workflow, I had a variety of NURBs curves, since the plan was to use them to create an accurate surfaces. One half of the surface ended up being accurate, but the other half left me wanting for more control, therefore in the end I constructed the webbing using traditional polygonal techniques.
After it was done, a bit of cleaning work after merging points needed to be done, after which a bevel was applied to some ribbing to get nice pointier edges.
Following is a video of how I actually model these vaults (without the ribbing), showcasing the techniques I developed over the course of the week. For comparison, the first vault took me almost 3 days, and the one after less than an hour.
(VIDEO UPLOADING, RELEVANT PICTURE PLACEHOLDER)
(VIDEO UPLOADING, RELEVANT PICTURE PLACEHOLDER)
End of the week and plans
Rest of the week was filled with refining the anatomy on the character model in low polygon version, followed by subdivision and actual sculpting. After tweaking the proportions a bit I signed it off for preliminary rigging.
Plans for following week (12.2.-18.2.) are as following:
- Finish preliminary rigging of the character, finish sculpting the character besides the facial features.
- Develop the vaulting in the semi-circular area of the cathedral
- Finish vaulting (Perhaps add ribs as well)
- Start animating few shots (hopefully)
Signing off,
Decay
Comments
Post a Comment