Evolving Facial Capture – Integrating DI4D Technology into Clear Angle Studios

The Clear Angle team has always operated with a single directive, to provide high-fidelity scan data that mirrors the physical subject.  Capturing talent in 3D is only the start, the real challenge for studios lies in processing data into production-ready assets without bottlenecks.

Creating a facial rig that preserves the nuance of a subject’s real life expressions has required extensive work to accurately re-topologise and stabilise a set of FACS scans. Tracking HMC data accurately onto high resolution geometry is a similarly important but equally difficult process. Clear Angle Studios’ integration of DI4D facial capture expertise directly addresses these pain points. With DI4D’s technology embedded in our workflow, we’ve launched ‘TopoTrack’ to bridge the gap between capture and character-ready assets.

Integrating DI4D’s markerless 4D facial capture algorithms into our pipeline brings over two decades of high-fidelity capture and processing research to our established global capture network. Since 2003, the DI4D team has pioneered the performance capture techniques  behind digital doubles seen in projects ranging from Blade Runner 2049 and Love, Death & Robots, to the Call of Duty franchise. DI4D’s proprietary optical flow tracking has set a benchmark for analysing dense facial data and stabilising performance at sub-pixel level. This expertise now operates as part of Clear Angle Studios, with co-founder Dr Colin Urquhart joining as Strategic Advisor for product research and deployment.

Inside the TopoTrack Workflow

For both film and game studios, the gap between receiving a raw mesh and having a usable, animatable asset is often defined by time-consuming manual cleanup and stabilisation. 

TopoTrack functions as a pipeline extension designed to streamline this process. It bridges the gap between scan creation and early asset development, ensuring that the fidelity captured by our cameras is preserved all the way to the delivery of the asset. It provides a consistent world space, tracks the eyes and jaw accurately and creates a direct bridge to HMC and 4D motion capture control.

The value of this new workflow lies in how it combines hardware precision with software stability. Processing requires pristine input data, and our hardware foundation produces texture-rich, geometrically stable datasets.

Our workflow is built on:

Dorothy (Head Rig)
A photometric facial capture system utilising 54 Canon SLRs, 22 Sony A7R3s and 13 machine vision cameras to record high-resolution FACS sessions with dense texture detail.

X-Series (Full Body Rigs)
A 204-camera system delivering consistent full-body datasets in just 1.2 seconds.

High-Density Illumination & Geometry Capture
Ensuring clean surface information and minimal movement variance across sessions.

Crucially, the pipeline respects studios’ need for specific topology. We deliver a neutral scan, which a VFX or game development team wraps with their preferred base mesh; whether that is a custom studio topology or an approved standard like MetaHuman. This step ensures the partner studio defines the exact UV layout and internal geometry for eyes and teeth.

Once the re-topologised neutral is returned, the software proposition takes over. We apply DI4D’s proprietary algorithms to transfer that specific topology to all additional requested FACS shapes. The system utilises sub-pixel level tracking from multiple camera angles to stabilise the geometry for isolated deformations. Rather than relying on manual alignment, the software mathematically locks the mesh to the performance, validating tracking accuracy even on extreme expression shapes. 

The output is a set of assets with extremely stable UVs that transition seamlessly across shapes. Because the stabilisation is handled at the processing level, we can generate perfectly aligned 8K Object Space Normal maps, RGB-packed displacement maps and Specular maps that match the geometry frame-for-frame.

At its core, TopoTrack exists to protect performance and to equip game developers and VFX supervisors with a cleaner, more efficient starting block for character work. By stabilising data at the processing stage, we ensure captured detail is preserved in the final asset. As pipelines grow more complex, our focus remains on delivering the most precise and production-ready datasets possible.

TopoTrack is the first step in a deeper integration of capture and processing at Clear Angle Studios. By embedding DI4D’s research directly into our workflow, we are aligning hardware and software development under a single R&D strategy. Future advancements will continue to build on this foundation to support the evolving needs of film and game production.

For more information about TopoTrack, visit the webpage here, or click here to get in contact.