Eventhacker
DE EN
← Back to the blog
Eventhacker Lab · Performance meets 3D

From the dancefloor
to a digital juggler.

· Eventhacker · 3 min read

Paule, an LED staff and sixteen seconds of club footage. Our starting point for a character that could bring the same energy to the Psy-Night page.

Five illustrated stages: Paule with LED staff, pose sketch, gray 3D model, wireframe rig and colorful render concept.
AI-generated concept illustration. All five panels illustrate the intended workflow, including the video-like reference panel. They are not screenshots of completed pose tracking or an existing Blender model. Enlarge the concept image
16.13 s
Video duration
484
Extracted frames
30 fps
Frames per second
24
Selected reference frames
The idea

Give the movement a body.

The appeal of LED juggling comes from the whole performance: the posture, the hands guiding the rod and the arcs of light it leaves behind. We want to carry that interplay into an interactive web character.

The planned setup separates character animation, staff movement and lighting. The Juggler button on the Psy-Night page would start the performance; colors, speed and trail length could be adjusted independently.

The planned workflow

Five steps towards the avatar.

01

Video reference

A real performance provides the visual reference: body posture, arm movement and the shape of the LED staff.

02

Pose sketch

A proposed skeleton makes shoulders, elbows and hands easier to reason about. The illustrated joints are hypothetical, not recovered tracking data.

03

Basic 3D model

The performer and prop are built as separate objects in Blender. A stylized likeness is the first goal.

04

Mesh & rig

A skeleton controls the character mesh. Hand targets and a separate staff control make the choreography editable.

05

Render concept

Emissive surfaces and fading trails bring the movement to life. The illustration shows the intended look.

Extracted

The reference is ready.

  • Every frame with its presentation timestamp
  • Selected images for visual comparison
  • Brightness, image changes and candidate luminous regions as JSON/CSV

These signals include background lighting and occlusion. They are visual references, not measured joint movements.

Next step

Build and animate.

Next comes the Blender character, its rig and an editable movement loop. A GLB export would bring the animation into the browser, where Three.js can control playback and lighting.

The model has not been built yet. A single dark camera view cannot deliver an exact 3D reconstruction; the first version will be an artistic approximation of the performance.