Lion Fountain

I originally planned to do my lion fountain key piece first, however I needed to spend time researching how best to approach it. I now feel I have a good idea where to start so began with a base model in Maya which I will then take into ZBrush to sculpt in the details and bake a normal map from.

First stage was to create the low poly lion. I used reference images on planes to get the basic shape as accurate as possible. I then took this low poly mesh into ZBrush and began to sculpt the details on. I began by trying to make the edges look like stone using the flattern brush and used the dam standard and clay-buildup brushes on the face and feet. I then used a hair like alpha on the mane.

After adding all the facial details to the lion, I exported the high poly, dropped down the lowest subdivision and exported this as my low poly. This way I could make sure that all the geometry lined up correctly and hopefully would prevent baking issues,

The next stage was to bake the maps. I originally tried to bake the maps using XNormals but found strange artifacts appeared on the back legs. Unsure what was causing the issue I tried using Mayas transfer maps option. This worked well and all the detail was transferred to the low poly mesh. I created a normal and AO map which I then took into Unreal 4 along with the base fountain mesh.

I noticed the AO map was causing ugly seams on the edges of the model and that the normals could be crisper so the next stage will be to improve the textures and maps for this model. With alpha fast approaching and my own personal deadline of having everything modelled and in by week 6 I decided to move on from this area and focus on getting the map ready for alpha. I will return when all models are imported and apply better textures.

UVing and starting to put interior assets in Unreal

After modelling most of the interior assets, I began to put them into Unreal. This meant I could play around with the layout of each room. I wanted to make sure each room did not feel empty yet was not overcrowded. I also wanted to tell a story with the assets I included. I wanted to make the rooms feel inhabited and furnished in a Moorish style as if the palace was being used again. With more time I would have loved too add a lot of small assets on the dressing tables and on surfaces but for now I have just allowed time for base assets to describe the role of each room.

Archways in Unreal 4

I decided to get my archways finished and in Unreal fairly switfly in order to get the main patio area of the scene looking more life like. I focused on getting the normals on the patterned arch working. I decided to go with a simple pattern for the normals and just use the diffuse map for the highly detailed centre pieces. I feel this will make the scene read better and make the normals look less noisy like the original tests.

I added a very simple base colour and AO map to the sections in order to check the normals worked correctly. At first, when imported into Unreal, the normals were reversed to what was being viewed in Maya. In order to solve this problem I went back to the normal map in Photoshop and reversed the green channel in the map. This fixed the issue when the normals were plugged into the Unreal 4 shader,

Another key decision I made with these pillars and archways was to have them as seperate pieces. The pillars are seperate as then I can use them elsewhere as assets inside the rooms and it gives me more flexibility for changing any specific instances if I feel the scene needs more variation in the future.

Interior Assets

Due to being at home for a few days and unable to access ZBrush to complete the sculpting on my highpoly arches, I decided to at least get some work done. I began to focus on getting some of the interior assets done so it would make it quicker to place everyhting inside Unreal4.

When beginning to think about my interior assets I started to create a moodboard. Many of the images I found were of 19th century Moorish furniture but the style is still accurate as it was copied from the original 13th century style.
I discovered that the shapes of the furniture were quite simple and that the work would come in the fine detailing on the textures.

Beginning to Model

The first area I focused on modelling was the archways in the main patio area. This took me longer than expected as I encountered a number of problems.
I began by making a high poly version of each then took these into ZBrush to add all the extra details I could not recreate in Maya.

I encountered a problem early on. When I imported into ZBrush to add my alpha detail, the mesh was not subdividing correctly so the alpha was not displaying correctly. After some experimentation I found that adding the tris and quads in Maya before hand fixed these issues.

The next issue I came across was how to make the pattern the right size and repetition. I initially experimented with simply dragging the alpha over the area of the mesh but this made it hard to tile more and resize. I then experimented with the projection master.This allowed me to reposition, tile and generally have more control over where my alpha got placed. I also found that the alpha I had originally created had too much noise in it so the detail was not crisp enough. I then decided to use a more simple design.

Above: Drag rectangle and noisy alpha. Below: Projection master and simple alpha

My next stage will be to add all the details I want on the arches in ZBrush, create a low poly and bake the details down into a normal map that I can import into Unreal 4.

Blockout Stage and Planning

After deciding on where I wanted the project to go I began to create a simple blockout in Unreal 4. This would help get a good scale for the environment and make it easier to work out how a camera could move out the environment. I used the geometry brushes in Unreal 4 for the basic wall sections and created a few quick static meshes for other assets to represent the scene better. I will go back and improve these models later on and do high poly bakes.

After creating a blockout I made a quick mock-up of what a cinematic fly-through of the level may look like. I focused on key features such as the central fountain in the patio and other assets that told the story of what each room is.

Planning:
After creating the blockout I began to plan out the rest of my project. I created a simple asset list that included the minimum amount of assets I needed to produce. If I end up feeling the room are too empty or that I have time left over I will create more meshes and assets to bring the scene to life.

Main Patio Area:

·         Lion fountain – one complete lion, one slightly broken to make some variation

·         Modular sections of columns 

·         Spare column for corners and extras

·         Water channels

·         Foliage – bushes, trees?

·         Roofs

Living Room:

·         Sofas

·         Chairs

·         Cushions

·         Tables

·         Lamps

·         Central fountain/water pool

·         Shutters on windows

·         Chandelier

·         Plants

Bedroom:

·         Four-poster bed with drapes

·         Bedside table

·         Lamp

·         Ceiling light

·         Dressing table

·         Wardrobe

·         Chest of drawers

·         Shutters and drapes on windows

Other:

·         Sky out of windows

·         Terrain out of windows and over walls on patio

By calculating how long it will take me on average for each asset (and overestimating slightly on each) I then created a basic gantt chart. This shows that if I stick to the plan below I can be finished about a week before the deadline. I would like to stick to this plan as it means I will have a week to polish the level however I am aware that other modules, events and circumstances may get in the way.

I had originally created a few basic milestones for myself. I aimed to get all the assets modelled and UV'd by week 6 latest so I have time to base texture before the alpha milestone at week 8. After creating the gantt chart I realised that this may be unacheivable and a more realistic milestone would be week 7. I had also thought about having a milestone for week 10 of everything finished but this again was pushed back to week 11 on my gantt chart after looking at the hours required for each task. If I find that I get things done quicker than illustrated in this gantt chart I will return to my original milestones of week 6 and 10.