Character Design for Movement Projects
For this week’s assignment, our production team (Michael, Ryan, Galt, and I) decided to split into two projects: a gym workout archive and a football play simulation. Both projects will rely on motion capture using OptiTrack and character creation with MetaHuman in Unreal Engine, where we are using the Live Link Face app to scan our faces and generate realistic digital doubles.
Project 1: Gym Workout
The goal of the gym workout project is to record fundamental exercises that can be performed without heavy machines, since our mocap sessions will be limited to the studio space. Instead of treadmills or bench press racks, we are focusing on bodyweight and dumbbell-based movements such as squats, lunges, push-ups, planks, curls, presses, and burpees.
By capturing these exercises, I hope to build a digital reference library of essential gym movements, where the character’s design reflects both athletic realism and training precision. The absence of equipment actually sharpens the focus on body mechanics and form, making the data cleaner for retargeting to 3D characters.
For the character design, we are using the Live Link Face app to record facial data and then generating 3D facial models through the MetaHuman plugin in Unreal Engine. This workflow allows us to apply our own scanned faces directly onto a preset MetaHuman body model and rig, giving us a hybrid result: personalized facial likeness combined with a standard rigged body ready for motion retargeting.
Below are the lists of gym workouts that will be able to capture in the studio.
Upper Body
- Push-ups (standard, wide, diamond)
- Pull-ups or chin-ups (if a portable bar is available)
- Dumbbell curls, hammer curls
- Dumbbell shoulder press
- Lateral raises / front raise
- Dumbbell bench press (can be simulated lying on a mat/bench)
- Tricep dips (using a bench or box)
Lower Body
- Squats (bodyweight, jump squats, goblet squats with dumbbell)
- Lunges (forward, backward, side)
- Deadlifts (using dumbbells)
- Step-ups (with a box/platform)
Core
- Plank (and variations: side plank, plank with reach)
- Sit-ups / crunches
- Russian twists
- Leg raises
- Mountain climbers
- Burpees
Project 2: Football Play
Ryan’s project explores the dynamics of a football play. Here, we plan to design at least two MetaHuman football players, capturing their coordinated movements during passes, runs, or tackles. Additionally, we are experimenting with recording audience and fan reactions, which could introduce a layer of environmental and cultural context to the dataset.
This project highlights team-based choreography and interaction, contrasting with the more individualized focus of the gym workout. The football characters will wear stylized uniforms and gear, making their silhouettes instantly recognizable.
Questions to consider
1. How can your character’s design contribute to the narrative experience of your movement archive?
By scanning both myself (an amateur) and a professional trainer, the character design highlights a contrast in skill levels. Even though we will use preset MetaHuman rigs, the personalized faces keep the movements grounded in real identities. This dual-character design makes the narrative more meaningful: the archive doesn’t just capture exercises, it captures the embodied knowledge gap between expert and learner.
2. What challenges might arise when translating this movement into data?
Professional trainers often demonstrate precise, subtle posture adjustments (core engagement, joint alignment) that might be hard for motion capture to capture accurately, especially without resistance equipment. On the other hand, my amateur movements may include mistakes, wobbles, or incomplete ranges of motion, which can be harder to retarget cleanly but are also valuable as authentic data. The challenge will be to preserve these distinctions in the mocap pipeline.
3. How can you ensure that your character design respects the values and sensitivities of the movement you are exploring (and avoids stereotypes)?
Instead of exaggerating “fit” vs. “unfit” appearances, we focus on actual performance differences between myself and the trainer. Using preset MetaHuman bodies keeps things neutral, while the facial scans personalize the models. This approach avoids stereotypes by not reducing people to physical ideals; instead, it shows the authentic diversity of movement practice.
4. What ethnographic research methods could you use to gather information about the cultural nuances of your character’s design?
- Direct observation of the trainer’s form during exercises, compared with my own.
- Video recording analysis of professional workout tutorials to study micro-movements.
- Auto-ethnography: reflecting on my own body experience as I attempt the same exercises.
- Interviewing the trainer about their embodied knowledge — what cues they think about while performing each move.
5. What cultural, historical, or environmental contexts will your character be embedded in, and how will this influence their appearance and movements?
This project is situated in modern fitness culture, which values professional guidance while also embracing accessibility for amateurs. The professional’s precise form embodies discipline, training, and expertise, while my amateur movement reflects the everyday struggles and imperfections of learning. By embedding both characters in the same digital environment, the project mirrors the real-world gym context where trainers and clients often share space, reinforcing the cultural significance of mentorship and practice.