The Weight of Data: Compression, Compromise, and Digital Legibility

V. The Weight of Data: Compression, Compromise, and Digital Legibility

        Even the most rigorous field documentation is ultimately shaped by the limitations of digital platforms. The technical ceilings imposed by public-facing 3D hosting services like Sketchfab—including file size limits (500MB for free users, 5GB for enterprise), polygon caps, and reduced metadata support—force users into a cycle of compression, simplification, and loss.1 This creates a second layer of erasure: not in the act of capture, but in the act of sharing.

During this project, we encountered these constraints when uploading the Robinson Family Home scan. The original intention was to publish the full environmental context of the site, including adjacent family land, vegetation, and a neighboring shed that appeared in the original point cloud. However, Sketchfab's file size and rendering limits necessitated a pared-down version. Models had to be decimated, textures compressed, and key spatial relationships deleted to meet the platform’s caps.2 These omissions were not curatorial decisions—they were architectural losses enforced by platform architecture.

The stakes are not trivial. When a Gullah home appears on Sketchfab as a floating model, detached from the land it stands on, it suggests a portability and placelessness that runs counter to its lived significance. The model may look “complete” to a casual viewer, but it has been stripped of ecological, familial, and spatial context. What remains is a visual icon, not a cultural system.

This disconnect reflects deeper questions of data stewardship. Hosting platforms optimized for creative industries or game development are not designed to handle the scale, sensitivity, or relational metadata required for heritage preservation. They privilege compression over completeness, visual appeal over evidentiary integrity. As a result, even high-fidelity documentation tools are undermined when the datasets they generate must be truncated for public view.

This issue is not unique to Sketchfab. Cloud-based mobile apps like Polycam and Scaniverse also face storage bottlenecks, limited export formats, and unpredictable rendering across devices. Their interfaces are designed for usability, not longevity. And while they enable wide participation, they also raise questions of ownership, control, and long-term access to data.3

The result is a double bind. Community-driven projects are encouraged to digitize cultural sites for visibility, but the very platforms that offer visibility often enforce invisibility—of context, of scale, of nuance.

The consequence is not just technical; it is epistemological. When compression drives curation, we risk presenting cultural fragments as wholes, and visual representations as stand-ins for spatial relationships. In the absence of intentional hosting models, immersive preservation remains compromised not at the moment of scanning, but at the moment of showing.

A video demonstrating the processing process. Featuring the mesh of the Robinson Home captured by the Leica BLK.



The Robinson Home’s full context could only be shown by video. Press Play!

References

  1. Sketchfab, Plans and Pricing, https://sketchfab.com/plans
  2. Field notes, Robinson Family Home model preparation, April 2025.
  3. Polycam, Pricing, https://polycam.ai/pricing; Scaniverse, How to Use, https://scaniverse.com/