In the first class, we introduced ourselves to one another first, and got to know many friendly classmates and teachers. Afterwards, we previewed various basic animation exercises we would work on this semester, such as those related to bouncing balls, action overlap and weight shift. Meanwhile, our teacher explained some basic animation principles to us and recommended a book titled The Animator’s Survival Kit. I have purchased its e-book edition, which is very practical. Besides, Ting introduced us to the golden pose in animation and its significance in class. A good golden pose has a clear silhouette, enabling viewers to understand the action at a glance while being dynamic and appealing. Thus, my first assignment is to copy different golden poses.
In the first week of the unit Design for Animation, Narrative Structures & Film Language, the class mainly focused on understanding the purpose of the course and how research will be developed over time. Instead of choosing a final topic straight away, we were encouraged to think about possible research directions and what kinds of questions we are interested in exploring.
I started by reflecting on my own animation background and what I want to improve during this course. At this stage, I am particularly interested in how visual elements, such as lighting and environment, can influence storytelling and audience experience. This made me realise that film and animation language is not only about technique, but also about how meaning and emotion are communicated.
We also discussed how animation can be used to engage with broader social and cultural ideas through visual choices. This week helped me understand that research does not need to be fixed at the beginning. Developing clear questions and exploring ideas gradually is an important part of the process.
I added a lot of life-like small items to the original scene, such as scattered odds and ends and old newspapers. This was mainly to reduce the common “emptiness” of 3D scenes and make the space look more like a real, lived-in room. At the same time, I also put two character models into the scene. Their presence added a layer of narrative tension to this rather depressing Chinese-style apartment, instead of leaving it as just a static environment.
Camera and lighting were the parts I adjusted repeatedly the most this week. For the camera, I fine-tuned the Depth of Field again. I hoped to guide the audience’s eyes more clearly in key shots. I also adjusted the Motion Blur, so that there would be a bit more blur and trail when the camera moved, instead of the overly clean look of a virtual camera. For lighting, I added some local fill lights to the room. For example, I put a bit of rim light around the edges of the characters and furniture, and used the contrast between warm and cool tones to enhance the layers of the images.
This week, I made a simple rough-cut preview and exported one or two versions of the video. I arranged the shots in the order set out in the Week 5 storyboard first. After putting them together like this, I could clearly see if the transitions between camera animations were smooth, and whether there were any parts where the pacing felt off.
During the editing process, I also tried adding some ready-made external video footage as an experiment. These were mainly used in the flashback scenes that haven’t been rendered yet—for example, some imagery related to social pressure. This step was more about previewing the overall effect, to check if the final video would look too visually monotonous.
At the same time, I started trying to add music and sound effects to the footage. I compared the rough-cut video with music of different tempos. Through this process, I gradually found a suitable vibe: a sound atmosphere that feels somewhat depressing but also carries a touch of nostalgia.
For every shot, I directly set keyframes on the camera. I mainly adjusted the camera’s position, rotation and focal length. By doing this, I first got the overall narrative rhythm of the video right.
The storyboard included some “shaky” shots that feel more like memories. So I spent some time figuring out how to create camera shake in UE5. I tried using the Camera Shake Base blueprint, and also experimented with manually setting noisy keyframes in the sequence. This was to mimic the feel of real handheld filming, instead of using a perfectly steady virtual camera.
By the end of the week, all the key shots had a basic camera animation draft. There is still room for improvement in smoothness and detail control. But this moving preview already lets me see more clearly how the scene will look when it’s fully animated.
This week, I officially started building the scene in Unreal Engine 5, gradually putting together all the separate assets I had created in previous weeks into the same 3D space.
First, I set up the basic structure of the room, and then imported the textured models I had finished in Week 8 into the engine. To avoid slowing down the overall progress, I used white boxes as placeholders for some assets that hadn’t been refined yet. This was mainly to check the overall composition and proportions, making sure they matched the shots planned in the storyboard.
The task that took up most of my time this week was adjusting materials, especially the Roughness parameter. This setting has a huge impact on the look of Chinese furniture from the 2000s, like worn wooden lacquered surfaces or plastic-like finishes. If the value was too low, the furniture would look too new; if it was too high, it would appear dull. So I tweaked it several times, focusing on whether the light reflection on the surfaces looked natural.
As for lighting, I kept it simple at first. I added some basic light sources and used Lumen’s real-time effects to check the light and shadow balance in the scene. Even though the lighting was basic for now, it already gave a rough idea of the nostalgic atmosphere I was aiming for.
In-class Learning WEEK9
In the ninth week, we learned and practiced Post-processing Volumes. Serra shared a simpler method to adjust them with us, which is using LUTs (Lookup Tables). We can change the overall scene effect by importing pre-adjusted LUT files, which function similarly to filters. The key advantage is that adjustments can be made in external software like Photoshop, making the process easier without having to tweak numerous parameters repeatedly in the engine.
Modified Pre-LUT
Modified Post-LUT
Comparison of effects in UE and Photoshop after applying LUTs
In addition, we learned about rendering, focusing on exploring various practical methods to optimize rendering effects and enhance visual quality. For example, regarding console variable presets, Serra provided us with many useful ones, and we also tried them out in practice.
This week, I mainly did some basic work. I focused on handling models between 3D software and Substance Painter to get them ready for texturing.
I spent quite a lot of time unwrapping the UVs of all the models I had built before and standardizing them. The process was a bit mechanical, but it was necessary to make sure that textures wouldn’t stretch when I painted them in Substance Painter .
In SP, I created a universal vintage wooden furniture smart material based on the features of Chinese furniture from the 2000s. This material had the glossy wood finish that was common back then, and I also added some light surface stains to give the furniture a more retro look.
I didn’t rush to import the models into Unreal Engine this week. Instead, I concentrated on adjusting the materials and textures first. Using the same smart material for the basic furniture is faster, and it also ensures a consistent style across the whole scene.
In-class Learning WEEK8
In the eighth week, we explored lighting in Unreal Engine in more depth, focusing on different lighting methods such as baked lighting, dynamic (real-time) lighting, and indirect lighting. We also became familiar with several common light types in UE, including directional lights, point lights, spotlights, and skylights.
After that, we carried out practical exercises in a simple scene, where we adjusted different light sources such as skylights and spotlights to see how they affect the overall atmosphere. We also learned some useful shortcut keys for lighting, for example using Ctrl + L to quickly adjust the angle of the directional light to simulate changes in sunlight.
In addition, we learned how to set up dynamic lighting effects. For instance, applying IES light profiles can help create more realistic and natural lighting results in a scene.
After finishing the overall scene layout, I focused on making the main assets in the scene this week, including the birthday cake and coffee table.
As I had already planned all the shots during the storyboarding stage, I knew the coffee table and the things on it were the key part of the whole scene. The coffee table has an important close-up shot in the video. So when I was modeling it, I paid special attention to small geometric details and surface realism, to make sure the audience could fully feel the story’s atmosphere.
The birthday cake and candles were the most important parts for telling the story this week. I used the Curve Tools to carefully draw the cream patterns on the cake. I also looked at many old-style whipped cream cakes from the early 2000s, so that the cake could have a real retro look.
For the coffee table’s design, I chose a simple style with some retro features. I hoped it could fit naturally into the whole feeling of a Chinese apartment in the 2000s, and become a visual part that brings out nostalgic feelings.
In-class Learning WEEK7
In the 7th week’s class, we focused on learning how to use the built-in rigging tools in Unreal Engine. After mastering the basic operations of the tools, we first used them to create specific effects like death animations. Later on, following Serra’s guidance, we selected a hunting dog model as the object for hands-on practice. By building and adjusting a series of related nodes, we successfully achieved procedural quadrupedal movement. Afterwards, we also made targeted attempts to tweak various relevant parameters, so as to make both the movement posture and trajectory of the hunting dog look more natural and smooth.
1. Core Content: Period-Specific Asset Creation and Layout
1.Asset Modeling and Uniqueness: The core setting of my project is the interior of a Chinese apartment from the early 2000s. Since most commercially available asset libraries are dominated by Western or contemporary styles, it is nearly impossible to find furniture that captures the specific texture and “soul” of that era. Consequently, I have had to manually model and craft the majority of the core assets from scratch to ensure period accuracy.
2.Layout Logic and References: Designing the overall layout of the room proved to be the most significant challenge this week. To avoid the space feeling hollow or structurally illogical, I conducted extensive research, consulting a vast array of old photographs, authentic floor plans from the period, and interior design case studies of vintage apartments.
3.Personal Connection and Realism: During the design process, I drew heavily on my own memories of the apartments I lived in during my childhood. This memory-based recreation not only helped me determine a realistic flow for the room—such as the characteristic narrow hallways and cramped living areas—但 but also imbued the scene with a sense of “lived-in” realism. This authenticity is crucial for achieving the cinematic visual impact I am aiming for in the final render.
In-class Learning WEEK6
In the sixth week, we learned about physics simulation in the Unreal Engine and explored several methods to achieve it, including the simulation effect for chain-like suspension. By setting the parent-child relationships between different objects and adjusting relevant parameters, we restricted the objects from moving randomly in all directions, thus creating effects similar to hinges and chains. In addition, we mastered the methods of baking physics effects in UE5, which allows us to apply physics simulation results in the engine more conveniently. We also learned to create object destruction effects, and successfully shattered a cube into fragments during the class.
This week, my main focus was to further develop the storyboard based on the written version completed in Week 4. I started to draw the storyboard by hand and organise the overall timeline of the animation.
At this stage, I translated the written descriptions into visual compositions, including camera distance, the relationship between the character and objects in the frame, and the approximate duration of each shot. Drawing the storyboard helped me to understand the pacing of the story more clearly, instead of relying only on written ideas.
While sketching the storyboard, I marked an estimated duration for each shot, such as 2–3 seconds or 3–4 seconds, and began to separate slower moments from faster, more emotional sections. For the middle part of the animation, where the mood changes, I plan to use shorter shots and simple transitions, such as black frames or quick cuts, to create a tighter rhythm.
Through this process, I realised that some shots that worked in the written storyboard were not necessary when visualised. After drawing them out, a few shots felt repetitive or disrupted the pacing, so they were either removed or combined during the timeline planning stage. This helped me understand that storyboards are not only a storytelling tool, but also an important way to control pacing and workload early in the production process.
Overall, Week 5 helped me gain a clearer understanding of the structure of the project and prepared me for the next stage, which will focus more on animation and shot testing.
In-class Learning WEEK5
This week’s tutorial helped me further clarify the presentation method for my project. Serra advised that I don’t have to rely on specific characters to tell a story, as objects and spaces themselves can also serve as “characters”. By documenting the changes of the same scene at different times — such as the disappearance of food on the table before and after a party — human presence can be implied without directly depicting people. Serra also suggested that I treat the process of change as a visual study through repeated shooting or time-based footage. This discussion made me realize that I need to simplify the format and focus more on the relationship between change, absence and narrative.
