When we started preproduction on Project Gold, we first had to establish the main beats of the story and translate them into a visual format. From the initial pitch, it wasn’t fully clear what the environment and visual effects would look like. Therefore, the storyboarding phase was crucial to understanding the scope, style, and story beats that conveyed the narrative. As Jericca, the director, pitched the initial story with a beat reel, we had a rough baseline to start working on the visualisation.
The idea was to begin with the most concrete scenes first—those that were already well-established in the initial pitch. This sparked a conversation with the director to visualize her vision and ideas and translate them into moving images.
Before we decided to use storyboarding for Project Gold, the initial plan was to start immediately with 3D previs as a way to visualize the story from the beginning. This task was assigned to Hjalti Hjalmarsson, whose job was to create a previs pass for the opening of the story. In essence, it is similar to storyboarding, but in this case, he used 3D proxy assets to set up the shots for the opening scene. The great advantage of this method is that once the 3D temporary assets are created, the previs artist has significant flexibility to adjust cameras and block rough animation to tell the story. Iteration is much easier this way; however, setting up temporary assets and modeling environments on the spot can be quite labor-intensive.
When starting with a blank slate, it can feel daunting and it takes time to figure out the story, spatial proportions of the characters, environment, and assets. These questions can be addressed during the previs process when a firm story is established, which can then be transferred into modeling and environment design relatively easily. However, too much exploration and iteration can become complex due to the constraints and relationships within 3D. For example, when a pass of previs was ready for review, the boat design and size had changed during the process. As a result, camera angles had to be adjusted to fix the new composition.
For Project Gold, it wasn’t clear how much exploration was allowed during preproduction, and expectations were not aligned in this regard. More complex shots that required effects or facial expressions were also challenging to achieve with previs. Additionally, the story relied heavily on metaphors, which were not fully fleshed out visually. Therefore, the decision was made to conduct most of the exploration through 2D storyboarding.
3D previs (previsualization) offers dynamic visualization, allowing filmmakers to explore camera angles, lighting, and movement in a way that static storyboarding cannot. It enables real-time adjustments, providing flexibility to experiment with complex sequences and ensuring creative decisions align with technical constraints.
Previs also helps with spatial awareness, having different departments understand scale, depth, and choreography in a virtual 3D space. It enhances collaboration by producing a shared, detailed vision of the scene that is easier for all departments to interpret. While storyboarding remains valuable for initial concepts, 3D previs offers a more immersive and precise planning tool for modern, effects-driven productions.
Storyboarding is faster and more cost-effective than 3D previs, making it ideal for early-stage brainstorming and conceptualization. It allows creators to quickly visualize ideas and make iterative changes without the need for extensive technical resources. Storyboards are simpler to produce, requiring only basic drawing skills or software, and can communicate the narrative flow and key visual elements efficiently. They are particularly useful for projects where detailed technical visualization isn’t necessary, such as smaller productions or straightforward scenes. Additionally, storyboarding retains a level of creative spontaneity, encouraging fluid exploration of ideas before committing to the more resource-intensive 3D previs stage.
The first scene ready for a storyboard pass was the one where Mikassa, the main character, falls through a false bottom and slides down into the abyss — a dark place at the very bottom, swallowed by a black liquid that we called Despair. From the concept art, there were some visuals showing how Mikassa would break apart and how the Despair might look, but how the scene would progress was left to the storyboarding phase to establish. As an animator, I quickly started filling in the in-between drawings and presenting the first draft as an animatic. Given the complex evolution of the environment Mikassa was falling through — featuring black liquid, cracks, and gold — it was important for me to convey a sense of speed, spatial transition, and the physical changes occurring to Mikassa. Additionally, it was essential to visualize the motion and character of the black liquid.
With some 3D proxy assets already created, including Mikassa, I aimed to use the 3D model as a drawing reference. This allowed me to navigate through the 3D space while adding Grease Pencil details on top of it. Initially, this seemed like a good approach, but when incorporating feedback, it quickly became apparent that this setup was too complex and not flexible enough to address feedback as quickly as traditional 2D drawings.
I soon realized that notes from the director had significant consequences for my carefully crafted animations, so it was crucial to present a rough draft first and avoid animating before the main beats were approved. It became clear that this initial pass was an exploration phase, focused on quick iteration and rough ideas that could be refined later. Although it was tempting to start animating right away—since that’s what I’m most comfortable with—it was too early to implement that level of detail in the storyboards.
It took some time to transition from an animation process to a storyboard process, where ideas are kept much looser and more flexible. To work more efficiently, I began showing rough thumbnails first — simple sketches that outlined only the essential story beats. These could even be stick-figure drawings, as long as they conveyed the necessary information and could be easily discussed with the director.
The key was to quickly bring an idea to the table and iterate just as quickly, addressing notes or exploring different angles for the same concept. Thinking of the drawings as disposable communication tools rather than finished artwork was essential for advancing the creative process.
Later in preproduction, it became clear that the opening, initially created with 3D previsualization, needed to be redone using storyboarding due to changes in the story. I discovered that starting with quick thumbnails and rapidly generating ideas was the best way to initiate a conversation with the director. Thumbnails, accompanied by brief written notes to indicate motion or clarify the intention behind each drawing, served as an effective starting point.
Frequent conversations with Jericca were essential to understand her reasoning behind decisions and distill them into actionable feedback. Since this phase was focused on collaboratively shaping the story, it was crucial to clearly communicate the intentions behind each beat, shot, and composition. Once these elements were approved after a few iterations, they could be further refined with in-between drawings and more detail, eventually being assembled into an animatic — a movie-like format where drawings are sequenced chronologically to convey the story.
This is the first time I was involved in the storyboarding phase of a project at the studio, but I kind of approached it like I do when animating a shot/sequence: I wanted to have the feedback from the director as soon as possible to check if I’m in the right direction. Thus for me what made sense was to start to do thumbnails, they were really rough and lacking detail, but it was a fast and disposable way to try different cameras or ideas in a really short time. When the director liked them, then I started to clean them up a bit, just to make it more comprehensive.
After all the feedback was addressed and Jericca was happy with them, I moved to an animated version of them, first just translating the thumbnails, and adding the poses I needed to convey the actions happening in the shots, again really rough. I decided to separate the characters in different layers and also the background in another one; this gave me a lot of control when I had to animate them. When the director was happy I continued, as before, cleaning all the drawings and making them look a bit more like Mikassa.
All this process was rather easy - making things in 2D in the first steps is easy for me, because you don’t need to care much about proportions and likeness of the model. However this created some problems later when I had to make the 3D layout, some of the camera choices I did on the storyboard were difficult to replicate because of the proportions of the different characters and I had to adjust them in 3D. Nevertheless, I think it was super useful to make them in 2D first, because it saved a lot of time.
For the storyboards, I used the Story Pencil addon, available on the extensions platform. It worked really good to keep all in the same file, and it was simple enough to understand. The thing that took a bit longer was to set all up the way I wanted, but once done, it was pretty ok to use during all the process. However, it becomes a bit cumbersome when using this with a team and managing iterations.
We also used some provisional rigs for some of the more complex movements. This helped to speed up the process of drawing, mostly when we translated the thumbnails to actual storyboards and we needed to animate them.
Since the initial concept art was quite abstract, it became partly the role of storyboarding and previsualization to translate these metaphors into visuals that supported the story beats. Visualizing an abstract world was challenging, but the storyboards played a key role in guiding the look development in the right direction. For example, seeing elements like the black liquid effect or the cracks in motion provided valuable input for both the concept and technical artists. It helped define the scope of what could realistically be achieved within the project’s time and budget constraints.
As a result, decisions were made — such as cutting the black liquid effect — to ensure the movie could be completed on time while preserving the essence of the story. Storyboarding also helped clarify ideas that might otherwise be interpreted differently by different team members. Seeing these ideas in the context of the story provided a shared understanding that moved the project forward.
Another significant question to address was how Mikassa would shatter and reassemble during her time in the abyss. Once the initial storyboards were presented, they facilitated decision-making within the art department, helping to finalize conclusions about her appearance, design, and visual evolution.
Seeing the storyboard in motion provides valuable insights into the pacing, atmosphere, and visual progression of the story. Once the initial pass of the boards is approved, they are typically handed over to editorial to be assembled into the film. Optionally, sound effects and music are added to enhance the experience and atmosphere. At this stage, the timing of individual beats can be fine-tuned, and shot markings can be distilled from the animatic.
Occasionally, the animatic would inspire new ideas for improving the progression of a sequence. In such cases, boards might be altered, added, or removed to strengthen the overall story.
After the animatic is approved, the boards are translated into shots through 3D layout. This phase involves preparing the shots for production by incorporating production assets and setting up the camera to finalize the composition and framing.
Join to leave a comment.