Concept & Idea Development
For this project, my group (consisting of Aveen Fitzpatrick, Odhran McManus, Dylan Short, Alexander Johnson and myself) chose the ‘tavern’ prompt. We then started to brainstorm ideas for the mood we wanted to capture, and the style we wanted our tavern to portray.
We started out with our asset list, and assigned each team member 4 assets;
The assets I chose were the bannisters, staircase, armchairs and window(s), so I began gathering references and sketching concept designs;
We didn’t want to use too much of our environment to purely house the staircase, since we had several other assets to incorporate, so in order to make it more compact, I suggested an L-shape for the design, which my team approved. I came up with three possible bannister concepts after researching their shapes, as I wasn’t sure whether I’d settle on a wooden or iron finish. I thought the third concept would work well for an iron finish being the slimmer of the three, but upon discussing this with my team, I decided that a mixture of concepts one and three might work well for a wood finish, which would compliment the aesthetic of our tavern as a whole. The mood we want to capture is that of a fantasy tavern, the likes of ones my team and I have seen in much-loved films from our childhood, such as Tangled, Shrek, and animated Disney classics such as Beauty and the Beast.
After discussing our sources of inspiration, we worked together to make a mood board (Aveen in particular was very active in this role in terms of presentation) of movie scenes, furniture references and colour schemes that we want to incorporate in our asset design as well as our final render;
After designing the staircase and bannisters, I then went to websites such as Wayfarer, IKEA, and Beliani to study armchairs and chaises-longues, which helped me to research and develop designs for my armchairs;
Originally, I envisioned the sofas and armchairs for the tavern to be quilted velour or leather, with a chaise longue reflecting the some design elements of the other armchairs, but still differentiating in style and shape slightly, to give the tavern a more ‘mismatched’ feel, such as you might see in an Irish pub.
Modelling & Sculpting
After my research and development, I began to sculpt a model for the chaise longue on Blender;
To block out the chaise longue, I used sphere primitives, which I then smoothed together in order to create the appearance of plump, quilted leather. I felt that this worked quite well on both the chaise longue and the armchair, which I sculpted next;
I loved the appearance I was able to convey with the smoothed spheres and crease tool, but upon working in the 2nd year room, I got talking to a final year student who suggested that retopology for these assets would be extremely time consuming, and it would be a difficult mesh to simplify, because I had used and combined multiple spheres, that was adding more faces to the mesh causing the computer to work harder, as opposed to one sculpted face. After importing to Maya, I started to share these concerns and, admittedly, I began to feel overwhelmed with the amount of troubleshooting I may face.
I emailed my lecturer Henry asking for his opinion on where to go from here, and he suggested I may have over-complicated things a bit with my model, so he provided a tutorial video for me which provided me with some new methods for re-building a simpler model. He introduced me to Maya’s ‘deform’ menu and the ‘blend shape’ setting, which allowed me to combine two separate meshes with a curve in order to connect the back of the sofa to the arm. I then was able to mirror this shape and combine the meshes to create the other side of the sofa. From a low poly model he used as an example, I was able to reverse-engineer how to unwrap the UVs of my new model in order to allow for more accuracy in regards to where on the model I place my textures, as well as making the most use of the space on my UV tile;
This experience has taught me that I need to be more active in seeking help when I need it, because truthfully, I felt that I was struggling to tackle a problem that I felt I should have found simple to solve. In order to improve my problem solving skills with regards to UV unwrapping, I intend to create some models of my own to unwrap and texture over the winter break in the hope that I have a better understanding of the best practice to do so when I return at the end of January.
Texturing & UV Maps
Below are the textures and mesh maps for my assets created in Substance Painter;
If I were to do this project again, I would like to be a little more adventurous in the texturing process by experimenting with things like texture masks and smart materials, but again, this is something I aim to practise over mid-term. I did however experiment with adding a photo texture to a material in Substance Painter for a stained glass window. I had originally used Photoshop to apply a mosaic filter to an ombre blue background in order to create the impression that it was composed of stained glass tiling, but upon discussion with my team, we used an image of a more ornate stained glass window which I converted into a photoshop texture to add some detail to our tavern environment;
Applying our textures to Unreal Engine proved to be a lengthy process. My textures were imported as were Aveen’s (one of my team members), however, our other team members had not finished all of their textures at that point, and once they did, there were some errors in our fbx file where multiple materials shared the same Lambert, so we had to keep exporting and re-importing the textures. I think towards the end of our project our time management and team communication could have been better. Some weeks prior I had suggested we all meet up to texture our assets in Unreal, but some team members felt it would be better to just texture individually and import those textures to OneDrive, after which point we can import them into the one file on Unreal Engine. This didn’t leave us with much time for troubleshooting in terms of materials being combined while texturing, and while I had done as much as I could with my assets up until that point, and I couldn’t add any lighting or animations until everyone else had their textures uploaded, which left me feeling very stressed.
Lighting & Cinematic
Our team ran in to some trouble initially while lighting our scene as we were unnecessarily trying to bake lights into our scene, so once we realised the problem, it was a process I feel we managed well. We used some animated planes to add moving PNGs to the chandelier assets to create candles and a flame in the fireplace, as well as creating rippling liquid spheres to surround my cauldron asset, which I though gave quite a mystical effect. The general lighting for the scene was primarily a warm white or pale orange to give our tavern a feeling of comfort and warmth. After having to wait for everyone on the team to catch up to a certain point however, I felt that I didn’t have as much time to experiment with the lighting and animation planes as I would have liked to, but I consider this a learning experience for my next assignment(s).
I also made some storyboard thumbnails to help me visualise what kind of shots I would like to establish in my cinematic render. The purple arrows indicate camera movement and the blue arrows indicate light sources;
Here’s how the shots compared to the storyboard frames;
When it came to the final render sequence, you can see there was some differentiation between what I intended for these shots, but I feel that the storyboard provided a useful guide.
For the background music, I wanted the tavern cinematic to have a medieval but familiar feel, which my team and I tried to reflect in our design of the environment. I found a page on YouTube where modern songs are performed using instruments specific to the medieval genre, such as the mandolin or lute. The song I chose was a medieval cover of Karma Police by Radiohead as it is recognisable but also isn’t overly upbeat, which again contributes to the relaxed atmosphere of the tavern.
I will be uploading a final version of the cinematic to OneDrive which I will then send to my lecturers along with my FBX files.
Reflection on Project
I think this assignment has provided several learning opportunities. It has shown me how important it is to maintain good communication with your team members, because it provides a foundation for how well you collaborate with each other in the later stages of a project. It has been difficult getting used to meeting up frequently and in person especially after the pandemic having prevented that for the last few years, but this project has provided a wake-up call on how important face-to-face collaboration really is. I’ve also learned that time management is something that may not be one of my strongest points, and I aim to plan my coursework and group responsibilities week by week in the future in order to help me stay focused and motivated. I also feel that I need to work on how I layout and unwrap my UVs, as I mentioned briefly earlier in the post. I intend to spend some of my free time over the break revisiting lecture content from first year in order to improve my confidence in this process, because I’ve found throughout this project, if I’m not entirely confident, that affects my productivity and general focus.