VES On-Set Data — Creator → Dataset → Consumer

How to use this

This Sankey diagram shows how on-set data flows through a film production: Creators on the left (the roles or departments that generate the data), Datasets in the middle (the data types from the VES Guide), and Consumers on the right (the roles or departments that use the data downstream). Band thickness reflects the number of dataset relationships.

Filters

Category — narrows to one of the 16 dataset sections from the Guide (camera & lens, lighting, scans, mocap, LED volume, etc.).
VFX Type — narrows to datasets tagged for a particular kind of VFX work (Basic 2D, Basic 3D, Character, Matte Painting, Complex, Mocap, Virtual Production, Real-time).
Essential only — restricts to datasets tagged with a small number of VFX types (default: 4 or fewer). The threshold is adjustable with the +/- controls. Intent: surface datasets specifically critical to a work type rather than the broader foundational stack.
Side — applies only when you hover or click a node in the chart. "Both" shows the full pipeline; "Creators only" or "Consumers only" shows just one direction. The category and VFX type chips at the top are not affected by this control.
View — Readable mode prioritizes label legibility (with vertical scroll). Fit to screen compresses everything to one viewport.
Zoom — +/- controls or keyboard +, −, 0.

Interaction

Hover any node for a preview highlight of its connections.
Click a node to lock the highlight (gold ring + underline). Persists through scroll, zoom, and density changes. Click the same node again or empty space to release.
Hovering a creator or consumer with Side set to Both walks the full pipeline arm. With "Creators only" or "Consumers only" you get just the direct datasets.

An honest note about the VFX Type tagging

The VFXTypes field in the source data is inclusive: each dataset is tagged with every VFX type it could be useful for, not those exclusively essential to it. As a result, types like Basic 2D, Basic 3D, Matte Painting, Complex, VP, and Real-time are co-tagged on most foundational data and read as "broadly applicable" rather than as a complexity tier. Mocap and Character are the genuinely narrow tags.

The Essential only toggle is a downstream interpretation layer that approximates "specialization" using tag count. At threshold ≤4, Matte Painting and Basic 2D return zero — that's the honest answer; these disciplines don't have data exclusive to them, they ride the shared foundational stack. The empty-state message in the viz says so explicitly.

Credits

Source data: VES On-Set VFX Data Collection and Usage Guide v1.0.0 by Sheena Duggal, with contributions from Sam Richards, Jim Geduldick, and Jake Morrison; technical support from Jean-Francois Panisset. Created for the VES Technology Committee.