Creating a Volumetric Performance Stage

For her VR performance “Suga’,” Movement Artist Valencia James wanted to use Cyark’s public LIDAR scan of the Annaberg Sugar Plantation as a stage. To help her realize her vision, we needed a way to simplify 25.1GB of scan data to quickly load and render on a range of browsers.

We found that many tools for handling point cloud datasets like Meshlab or Cloudcompare ran out of memory while decimating such a large scan. We found that by using PDAL, we could efficiently reduce the number of points in the pointcloud. This gave us the first steps in creating Valencia’s virtual stage. We then treated the data to best fit the needs of the performance, creating versions of the mill that used solid meshes and point clouds.

We were able to create a visually distinct stage for “Suga” that fit the conceptual framework and technical constraints of the performance. We were able to reduce the isolated stage to 6.1mb as points and 2.2mb as a mesh for quick downloading and rendering in Mozilla Hubs.