This week focused on establishing the technical foundations of the project, beginning the production of the first environment as well as establishing a reliable workflow between C4D and UE5; and continue to develo the teddy bear’s interaction system.
One of the first tasks was setting up and testing the workflow between Cinema 4D and Unreal Engine 5. As the final film will be rendered in Unreal Engine, I explored the available transfer methods between both applications and began preparing a pipeline that will allow models, animations, and scene data to move efficiently between the two pieces of software. Establishing this workflow early is important to avoid technical issues later in production and to ensure that assets created in Cinema 4D can be integrated into Unreal Engine without significant rework. During this process, I also identified the need to make objects editable (“Current State to Object”) before export. This is necessary both to prevent simulated elements from continuing to move unintentionally and because Unreal Engine does not support Cinema 4D’s procedural objects.
Alongside this pipeline development, I started building the first environment of the film: the childhood blanket fort. Rather than relying on cloth simulations, I chose to construct the fort using a simple supporting structure that would later remain invisible in the final scene. A plane object was then positioned over this framework and manually shaped to create the folds and overall silhouette of the blankets.
This approach provided several advantages. It allowed greater control over the final composition, avoided the unpredictability of cloth simulations, and produced a cleaner stylised result that better matches the visual language established during pre-production. The goal was not to achieve photorealistic fabric behaviour but to create an environment that feels believable while maintaining a simplified and nostalgic aesthetic.
In parallel, I also experimented with cloth simulation using pillows as a test case, inspired by a reference demonstrating how soft objects interact and deform under pressure. These tests helped me better understand the behaviour of cloth and soft-body systems, even though the results reinforced my decision to prioritise manual modelling for the final environment due to greater control and consistency.
A large portion of the week was also dedicated to testing deformation methods for the teddy bear. Since physical interaction with the toy forms a key part of the narrative, finding an effective way to represent hugs, compression, and touch is essential. Initial experiments focused on Soft Body simulations combined with collider objects. These tests successfully generated squash and compression effects but often produced unstable results, unwanted secondary deformations, and a lack of precise control over the final shape. While physically accurate, the behaviour was difficult to art direct and did not consistently produce the visual result required for the film.
Following these tests, I began exploring Pose Morph as an alternative solution. Instead of relying on simulation, Pose Morph allows specific deformations to be sculpted and animated manually. Early experiments suggest this method offers significantly more control over the teddy bear’s shape during moments of interaction while still communicating softness and physical weight. This approach is currently proving more suitable for the project’s needs, particularly given the importance of maintaining the bear’s recognisable silhouette throughout the animation.
Initial animation tests focused on recreating the sensation of a hug through controlled compression of the teddy bear’s body, arms, and head. Particular attention was given to the timing of the interaction, including a short hold at the peak of the embrace before the bear gradually returns to its original shape. This pause proved important in reinforcing the emotional quality of the gesture, allowing the moment to feel intentional and reflective rather than purely mechanical. The resulting deformation retains the character’s identity while effectively conveying comfort, attachment, and physical contact.
Here’s a comparison from the teddy worked on last week and this one.
Last week:
This week:
During the transfer process from Cinema 4D to Unreal Engine 5, an important technical challenge emerged. The teddy bear’s hug animation relied on two different deformation systems: skeletal animation driven by an armature for the limbs and head, and a Pose Morph deformation used to compress the bear’s torso during the hug.
While FBX export successfully transferred the skeletal animation, the Pose Morph deformation was not preserved correctly inside Unreal. To overcome this limitation, both animation systems were baked into Alembic caches. The torso compression was exported as a geometry cache, while the armature-driven movement was exported separately. These cached animations were then combined inside Unreal Engine, allowing the full hug deformation to be reproduced accurately.
This workflow provided a more reliable solution than relying solely on FBX export and ensured that both the physical movement and the subtle memory-like compression of the teddy bear were maintained.
Overall, this week represented an important step from planning into production. The first environment has begun taking shape in 3D, the Cinema 4D to Unreal Engine workflow has been established, and technical experimentation has helped identify a more controllable approach for the teddy bear’s deformation system. These developments provide a stronger foundation for the next stage of modelling, animation, and scene assembly.
The scene 1 is successfully imported: