When our friends at Hahn first approached us with the Close Call Support Group concept for North East Texas Mobility Authority (NETRMA for short), we quickly identified it as an opportunity to put our hybrid 3D + AI production workflow to the test. The workflow combines our traditional 3D pipeline with the use of AI to rapidly generate concept art, character design, look development, and as a means of rendering, while storytelling and animation retains a human touch.
Each stage of the process leverages the strengths of the technology involved, resulting in a workflow that is both efficient and cost-effective, while maintaining full artistic control and integrity. A key consideration when working with AI is the ability to refine and iterate; the process must feel firmly directed by the creator. This approach allows for precise control, enabling even subtle adjustments to visuals and performance.
Using the above character artwork generated by the agency as a foundation, we expanded on the designs and developed detailed character sheets for production. We also introduced subtle props and signs of wear to each character to enhance their individuality and implied backstory.
Originally, each :30 second spot was intended to feature only two characters seated side by side as seen in the original character artwork. Working together with the creative team at Hahn, we evolved this into a broader concept by bringing all characters together as part of a unified support group across the campaign. We also extended the limited background environments into larger, more flexible spaces which could be viewed from multiple angles. Look development stills were created to explore and refine the overall tone of the campaign.
The modeling and rigging process followed a largely conventional workflow. Characters were modeled and rigged in Cinema 4D, with relatively simple facial and body rigs. This supported a stop-motion-inspired aesthetic, with restrained movement—giving the impression that each pose was carefully arranged by hand.
Once camera angles were matched to the concept art, we built simplified gymnasium geometry from multiple perspectives and stitched these together into a cohesive environment. AI-generated environment imagery was then projected onto this geometry to create a flexible, photorealistic backdrop for our woodland friends.
Animation was approached using a traditional 3D workflow. From the outset, we were clear that camera work and character performance should remain under direct human control, rather than being delegated to AI. As a result, all animation was crafted by hand as seen in the animatic below.
We then output multiple render passes—including viewport, line, and depth passes—to guide the AI during the final rendering stage.
Rendering is where the hybrid approach fully comes together. While powered by Ai, the process mirrors a traditional rendering workflow. We generated starting frames using character sheets and line work exported from Cinema 4D. The starting frame, along with animated render passes were ingested into the Ai system to produce the final imagery by applying the established look across the full sequence of shots for each spot in the campaign.
The spots are finished with color grading along with sound design, and mix from our friends at Auralation Studio.
Want to see if our hybrid 3D + AI workflow is right for your next project? Get in touch!
