OpenVolCap

What is OpenVolCap?

OpenVolCap is a free and open source volumetric capture system for capturing 4-dimensional human performance data. Using DIY Raspberry Pi-based cameras, the “Chump” system allows for simple, affordable “one-click” capture and data processing pipeline to Postshot and Instant NGP formats. OpenVolCAp allows for easy training with the light field optimization method of your choice, whether it be gaussian splatting, neural radiance fields, or any other volume rendering technique.

Example output from our 8 camera OpenVolCap setup.

Intuitive User Interface

Simple UI OpenVolCap is operated using a single 8-panel GUI, which serves as a one-stop shop for camera control, intrinsic and marker calibration, and dataset preparation.

DIY Cameras

Image of Chumper in Stereo Configuration

Raspberry Pi-based cameras Chumps uses Raspberry Pi-based cameras in 3D printed enclosures to allow for 24 or 30 fps synchronized PoE capture systems at an affordable price. The chump system is modular and upgradable, and can be built in mono, stereo, and RGB-D configurations.

Automatic Calibration

Apriltag Calibration Instead of relying on structure from motion techniques like COLMAP, Chumps uses Apriltags to allow for automatic per-capture or per-frame camera calibration, allowing for faster setup, and sparse camera arrays.

Configurable Image Processing

Automatic Processing With a single click, prepare volume datasets into Postshot or Instant NGP formats with optional subject masking, as well as stereo and monocular depth pipelines.

Pointcloud and Preview Mesh Sequence Creation

Geometry Creation Chumps creates initialization point clouds using stereo projection and voxel carving methods. Additionally, chumps will generate a blender scene with preview mesh sequences for verification and previsualization.

3D Printed Enclosure ("Chumper")

The node case is 3D printed, housing the Raspberry Pi 5, PoE hat, and swappable camera lid configurations.

Drag to rotate • Scroll to zoom

Check out our proof of concept film!

Here is a short film we made as a proof of concept for our system, enjoy!

Citation

@misc{montag2026opensource,
  title={An Open Source Volumetric Capture System for Production Applications},
  author={Montag, Aidan and Phillips, Flip},
  year={2026},
  howpublished={Capstone Project Report, Rochester Institute of Technology},
  institution={Rochester Institute of Technology, Motion Picture Science}
}