Our Technologies
SCROLL DOWN
↓
Cross polarized 3D scan
Over the past few years, we have consistently improved our scanning rig. We started with a 60-camera setup and have grown to a state-of-the-art 192-camera configuration.
However, the most significant innovation we’ve introduced is the transition from flash lighting to fully customizable and controllable LED lighting.
This advanced lighting setup allows us to capture extremely fast image sequences using both parallel and cross-polarized filters.
As a result, we are now able to extract from our raw data:
– True albedo color, free of baked-in lighting,
– And a specular map that enables more accurate relighting of the model and enhanced extraction of fine skin details.
OLAT CAPTURE
Our 360° lighting setup, combined with our Canon R5 cameras, allows us to capture OLAT (One Light At a Time) references. These captures are extremely valuable for building a precise and accurate lookdev scene.
We provide our clients with the aligned face scan, the camera positions, and the corresponding 41 HDRI maps, all carefully calibrated for consistency.
With each lighting condition, you can assess your results against solid ground truth data from four different points of view, ensuring both reliability and realism in your workflow.
SYMBIOTE PARTNERSHIP
With our latest developments on our 3D scan, we can now leverage cross-polarized and OLAT captures to offer more tools to our clients for creating believable digital humans.
Thanks to our partnership with TexturingXYZ, our clients can receive enhanced skin textures and all the maps required for look dev – powered by Symbiote’s cutting-edge technologies: Hyperskin 2.0 and Hyperlook X.
GAUSSIAN SPLATTING
Our setup is no longer limited to producing 3D meshes. We can now use captured data to generate incredibly realistic Gaussian splats.
This technology allows users to achieve highly photorealistic results with significantly less effort. Since Gaussian splats are inherently photorealistic, they offer an excellent solution for product visualization or for integrating realistic elements into 3D environments.
As a final benefit, these splats can be optimized for web visualization, offering impressive visual quality with a lightweight footprint.
VIDEO CAPTURE
Our capture system can be used to record video data, either for seated 4D capture or to acquire ground truth data for training accurate face data used in ML algorithms.
Our fully customizable lighting setup allows us to recreate any lighting scenario or pattern, making it easy to match any capture protocol with precision.
As a final touch, we can use RGB filters to project any color onto the subject, and even display HDRI images directly on the lighting rig, enabling a wide range of creative and technical possibilities.
OBJECT SCAN
Over the years, we have continuously improved our object scanning pipeline.
We can now leverage photometric and laser capture techniques to accurately scan objects of various sizes.
Turntable
Photometric capture with Canon R5.
Laser scan
We use a Shining 3D Combo scanner for tough or complex surfaces.
The turntable
Leveraging our programmable lighting setup and a turntable module, we can precisely capture a wide range of objects, from shoes and bags to statues. As long as the object fits within our rig, we are ready to give it a try.
Our capture process allows us to extract albedo, specular, roughness, and surface maps directly from the data, ensuring high-quality assets suitable for lookdev and production.
The turntable
Leveraging our programmable lighting setup and a turntable module, we can precisely capture a wide range of objects, from shoes and bags to statues. As long as the object fits within our rig, we are ready to give it a try.
Our capture process allows us to extract albedo, specular, roughness, and surface maps directly from the data, ensuring high-quality assets suitable for lookdev and production.
Laser scan
For highly reflective objects like leather goods or metallic items, we use a laser scanner to generate an initial layer of geometry. This is then perfectly aligned and combined with a photogrammetry-based capture.
This approach allows us to leverage the strengths of both techniques, resulting in highly accurate and visually optimized 3D models, even in the most challenging surface conditions.
Laser scan
For highly reflective objects like leather goods or metallic items, we use a laser scanner to generate an initial layer of geometry. This is then perfectly aligned and combined with a photogrammetry-based capture.
This approach allows us to leverage the strengths of both techniques, resulting in highly accurate and visually optimized 3D models, even in the most challenging surface conditions.