Modeling Maynooth Castle Part 5 – Woe to Materials

I’ve decided that Materials in 3DS Max are not my favourite thing.  In fact, I think they are my least favourite thing.  I’ve been playing around with them for the past couple of days, trying to get a stone material that looks just right.  While I’ve had some minor success with a material for the buildings, I have yet to make it look like individual stone work (currently, it looks like each building is carved from a single rough-hewn stone).

The Process So Far

As I mentioned in my earlier blog post, my original thought on materials was to use a multi-subobject approach.  I would apply diffuse and bump maps using images taken of the exterior of the keep, then use secondary materials for the roof.  This approach was largely unsuccessful. I ended up rebuilding the roofs on the various buildings as other objects, so I could apply separate materials to them.  For the stone work on the buildings, I created a blend material that used an image from the castle wall as the map for the blend.  This seemed to work better as it presented more of a rough stone look.  I then applied a UVW Map modifier to each object and applied the material.  With the exception of the material creating the illusion of a single stone instead of multiple stones, the material application worked rather well.  However, when applying the same materials to my wall objects (which also had a UVW Map), the material looks painted on and doesn’t have the same texture as the buidings.  See below for an example:


I’ve tried messing with the tile values on the UVW map for the wall objects, but that doesn’t seem to have the effect I was hoping for.  I may scrap the material all together and return to using images applied as bump, diffuse, and displacement maps in order to create a more individual stone look.  As for the walls, I suppose more trial and error is involved.  I will continue to work with the UVW map settings; I may also separate materials for the walls, so I can further control how they are applied.

Modeling Maynooth Castle Part 4 – Materials & Lighting

Since my last post about Modeling Maynooth Castle, my progress has been solid but largely uneventful.  I haven’t encountered any major issues or problems outside of the normal day to day frustrations of modeling, such as paying extra attention to that one polygon that doesn’t seem to want to be shaped the right way or fixing an accidental extrusion, etc.

After making significant progress on the model itself, I decided to start applying some materials and lighting.  The lighting itself was relatively simple.  Since I’m dealing with a castle that was destroyed in the mid 16th century, I decided I was only going to use natural light.  Thus I was able to easily implement a daylight system.  While I couldn’t quite set the date of the daylight system to 1534 (the year the castle was destroyed), I was able to take it back all the way to 1583, which is the farthest back the software would calculate. I set the time of day as noon and set the appropriate latitude and longitude in order to calculate the position of the sun.  3DS Max made all of this very simplistic, and I had no issues implementing the system.

The materials have proven to be a bit trickier.  I took pictures of the current keep walls, as I hoped to use those as materials.  I pulled one of these pictures into photoshop and used the patch tool to create a seamless, tileable texture.  I think applied this texture as both a bump map and a diffuse map to a material and mapped the material to the keep.  The results are below:


The result isn’t bad, but I don’t feel it’s anywhere near complete. I started experimenting by applying a UVW map.  Then I decided to scrap the original material and create a multi-sub-object material that would allow me to set different materials for different areas (such as the roof, the wood paneling that is part of the roof, etc).  I also decided to create blends and use the images of the castle walls as maps for the blend.  The result ended up with the castle walls themselves looking a little cleaner:


I think the walls of the castle itself look a little better (ignore the giant cross shape in the middle for now.  Those are actually separate objects I haven’t applied the new material to yet).  However, I don’t think the multi-subobject approach is going to work for the roof:



As you can see, not only does the material not look different (it should have more of a tiled slate look rather than stone) but you can also see each distinct polygon where the material was applied.  I think I’m going to have to trash the roof and rebuild it so it is a separate object that I can apply its own material to.

I’m going to continue working on it to see how I can improve upon it.

Modeling Maynooth Castle Part 3 – The Keep

I started modeling the keep a few days ago and felt I was making good progress.  Unfortunately, I was wrong.  However, I’ve learned a very important lesson:  when doing extrusions on an object, always zoom out to make sure you didn’t accidentally destroy the geometry.  Sadly, this lesson cost me about 3 hours’ worth of work.  Thankfully, I started making regular backups of my saves; otherwise this could have been much worse.

The Process

I started the keep with just a standard cube. I then constructed small towers that went along the tops, and for the crenellations, I did what I have always done—I extruded the polygons of the cube (after converting the cube to an editable polygon). I’ve never had a problem with this. Here is what the top part of the keep (which was the area I was focused on) looked like when I noticed my problem.
Top of Keep
I then zoomed out in order to inspect some aspects of the roof of the keep (for which I was about to create the pitched roof). That’s when I noticed my problem.
Messed Up Keep
As you can see from the photo, quite a bit of the geometry of the keep is distorted. Random sections are extruded or missing. The bottom of the keep was completely distorted. I probably could have fixed it, but I decided the amount of work it would have taken me to fix it most likely would have equaled the amount of time I spent getting to that point since my last back up. So I decided to cut my losses and simply revert to my last backup.

What Went Wrong?

I’m not entirely sure where everything went wrong, to be honest. The only thing I can suspect is that I accidentally selected other polygons while I was selecting the polygons for the roof (either that or I had failed to set the height segments when converting to an editable polygon and thus was extruding the entire height of the keep as opposed to one that was supposed to be closer to 1m3.

Lessons Learned

One lesson I’m taking away from this is to always check my line segments before converting to an editable polygon, and another is to frequently check the entire object when making modifications which affect the geometry. But the most important lesson here is to create FREQUENT backups of my scene. Thankfully, this really saved me this time (as I only lost a few hours’ worth of work), so it’s definitely a lesson I have taken (and am continuing to take) to heart.

Modeling Maynooth Castle Part 2 – Crafting the Walls

In my first blog on modeling Maynooth Castle, I discussed the goal of my final project for AFF621 and a little bit of the background regarding the castle itself.  As I began the process of modeling the castle, I decided the first place to start would be to model the walls, which is what I intend to discuss today.  However, before I get started, I’d like to take a few moments to discuss the planning of the model itself.

A Little Project Planning

Like some graphic design software, such as Adobe Photoshop, 3DS Max requires you to think through how the model is going to be constructed, so that you can be organise your scene.  In order to start this process, I started thinking about what kind of objects would be in my scene, and how I wanted to arrange them.  I came up with the following major items that would need to be a part of the scene:

  • the castle walls
  • the keep
  • the church
  • the gates
  • the living quarters / kitchens / etc.
  • miscellaneous buildings (such as the stables, barn, etc.)
  • the river(s)
  • misceallaneous nature (trees, grass, any other surrounding items, etc.)

Depending on how quickly I am able to execute on the model itself, I may or may not be able to complete everything, so I’ve decided to prioritise the first 5 items with the hope of adding the other objects should I have the bandwidth.

In order to facilitate this process, I decided I would keep each of the above mentioned-sections in their own layers.  This way I could easy turn off entire sections if needed (which could be helpful while fine-tuning other aspects of the scene). Additionally, by grouping them in layers, I can focus on each section at a time—very similar to how I would construct a physical model from legos or other building material(s). With these decisions made, I decided it was time to start the actual modeling process.

Constructing the Walls

I decided to start with the walls for two reasons:

  1. They would form the boundaries of the rest of the inner buildings of the castle itself
  2. They would most likely be the easiest thing to model

I decided to do a quick Google search to begin with just to see if anyone had modeled anything similar in 3DS Max and might therefore have some insight into the process of building walls.  I was lucky enough to happen upon a video that discussed how to construct a castle using 3DS Max.  I watched some of it and decided that it may be useful later, so I’ve filed it away for future use. I also found a script that was created to make the construction of walls easier.  I thought this might be the ideal script to facilitate the creation of walls, so I downloaded it and started using it in my scene.

Unfortunately, the script wasn’t quite what I was looking for.  It’s the kind of script that is great for creating brick walls, but one of the biggest problems I had is that, while it allowed me to set the number of segments for length and height, I couldn’t set the number of segments on the width.  Since I needed to create crenulations along the top of the walls, using this script would have meant manually creating individual crenulations as opposed to using some effects on the wall itself.  I also discovered the script was very computationally intensive and a single wall greatly reduced the processing power of my laptop.  Ultimately, I decided to discard the script and create my own walls from scratch.

I decided to start by creating a box that met the dimensions of the wall I was trying to construct.  I then gave the wall an appropriate number of segments for length, width, and height that would simulate brick work. For example, the south wall is 86.8 metres long, 3.51 metres wide, and 4.96 metres tall. As a result, I gave it 86 length segments, 3 width segments, and 5 height segments which makes each stone roughly 1 metre x 1 metre x 1 metre.  This was an arbitrary decision but one I felt was reasonable.

After drawing out the wall, I converted it to an editable polygon and then used the extrude tool to create crenulations along the top.  I decided each crenulation would be roughly 2 polygons wide and 1 polygon deep, and would be extruded by 1.016 metres.  I then built the remaining 3 walls following the same pattern.  When I was finished, the scene looked as below.  Note: each wall is color coded so I can easily keep track of what wall it is.  The south wall is red; the east wall is yellow; the north wall is green, and the west wall is blue.


Next, I needed to fit the walls together.  The castle walls themselves do not form a perfect square, but rather bend at some points to create a sort of rounded, circular square.  There are also some breaks in the walls where the walls either protrude or where the walls are broken up by gates or other buildings. I accomplished this in a number of ways: by creating generalised boxes of the appropriate size as placeholders, by cutting up the walls to create the protrusions, and by using soft selection to bend sections of the walls. The outcome can be seen below.


As you can see, there was a drastic change in the state of the walls from their first construction.  It’s slowly starting to come together.

Further Struggles

One of the things I’ve struggled with a bit is the application of materials. I have some images of the castle walls themselves. I was hoping to apply them as both a bump map and a diffuse map to create the illusion of stone on the walls.  Unfortunately, it hasn’t worked so well.  Either the images tile too much and look fake, or they blur and, well, look fake.  This will be something I’ll have to work on more later.  I’m sure I’ll blog about it in future posts.

Next Steps

Next, I’m going to focus in on the castle keep and the enclosing walls. Hopefully this will be a little easier than the outer walls, especially now that I’ve picked up a few tips and tricks. More to come later . . .

Modeling Maynooth Castle Part 1

For my final project for AFF621 – Remaking the Physical, I was tasked with creating a 3D model of a cultural heritage object.  After considerable deliberation, I selected Maynooth Castle, which is a castle that once stood in the heart of Maynooth and was a major seat of power in Ireland from its construction in the latter part of the 12th century until its seige and destruction in 1534.  I selected it for two primary reasons: my love of ruined castles and the lack of information showing what the castle likely looked like during the height of its power, prior to its destruction.

Over the course of several blogs, I will recount the steps I have taken to produce the model and what trials and tribulations I have endured.  This is my first forray into creating 3D models, and while I’m excited to see the final output, I must admit to some trepidation regarding the scope and ambition of my project.  But I have always risen to a challenge, and this time is no different.  Hopefully, my efforts will be met with success, and by documenting my process and trials as I go, hopefully not only will I learn something from the experience, but so, too, will my readers.

Researching the Castle

The castle itself was originally built in the latter part of the 11th century and was a seat of power for the Fitzgerald family.  While I will save much of the formal history of the castle for my official report, I will note that the castle stood as a whole for nearly 300 years.  It fell in 1534 after a 10-day seige by the British, thanks to the rebellion of Thomas Fitzgerald, 10th Earl of Kildare.  While the castle was restored in 1630, it was destroyed again in the 1640s during the Irish Confederate Wars.  Since then, the castle has remained in ruins and is now run as a cultural heritage site by the State. It is open during the summer months for tours.

Unfortunately, due to the age of the castle, not much information remains regarding what it may have looked like during the height of its power, so most of my model is based on speculation of architects and other historians much more qualified than I.  I will be building my model based on their work, and my report will detail those sources in depth.

Some of the information I received was from some older documents stored in the special collections area of the Maynooth University library.  These included a map of County Kildare (which had a lovely write-up regarding the history of the castle), as well as some hand-drawn images of the castle ruins at the time of publication in 1783.  Additionally, the library was also able to provide me with a development plan that was created for the Office of Public Works in the mid 1990s that included some architectural plans that were drawn up during the castle’s first reconstruction in 1630. Additionally, this document also contained some speculation from the architects creating the development plan as to what the castle may have looked like in the 15th-16th century prior to its fall.

Looking at the Models

Another source of information was the model housed within the castle.  When Maynooth Castle was converted into a cultural heritage site, a scale replica of what the castle likely looked like was created.  This scale model is housed within the keep (which serves as a type of museum for the castle itself).  Catherine O’Connor, the supervisor of the site, was gracious enough to provide me with early access to the castle keep (which is currently closed for the season) so that I could photograph the model and take measurements of each of the buildings.  This will allow me to construct my 3D model as accurately to this model as possible.  Ms. O’Connor was also able to provide me with some additional documents that detailed an archeaological excavation of Maynooth Castle that was conducted in June of 2000. Finally, Ms. O’Connor also provided me with contact information for one of the architects who worked on the development plan.  I will be reaching out to him over the next few days in hopes that he can provide me with any further information that may be of use.

Next Steps

The next thing I will begin tackling is the creation of the model itself.  I will likely begin by trying to create the outer walls, as well as the keep.  I will be doing this from the ground up in 3DS Max.  My next post will be written after I have begun tackling some of these aspects and will detail what struggles (and hopefully triumphs) have resulted from this endeavour. Stay tuned . . .