presetsone-click configs
01sourceupload or demo
🎞️
Drop a video here
or click to browse — mp4 · webm · mov (webm with alpha = pre-keyed)
idle
awaiting video…
02samplingframe density
subject grab — crop the subject out of each frame
transparent background — use the video's own alpha channel (pre-keyed footage)
Higher = cleaner subject and fewer stray “flyaway” pixels.
Importance sampling keeps the busiest frames when downsampling long videos.
1 frame kept per N source frames (assumes ~30fps).
Pixel brightness change that counts as "moving".
03opacityvoxel classification
Opacity of moving voxels — action, subjects, transitions.
Opacity of background voxels — kept low to avoid occlusion.
Brightness of each frame's boundary — reveals the volume faces.
Overall brightness of the additive glow — slices auto-scale, so lower this if the volume washes out.
04transformaffine shear
Panoramic synopsis for horizontal / panning camera moves.
Length of the time axis relative to frame height.
Orientation of the frame timeline through the cube.
reverse sequence (play front → back)
05deformspline skeleton
enable skeleton deformation
X ↔ time (top view) — drag points
Y ↔ time (side view) — drag points
A central Catmull-Rom spline runs along the time axis — warp the cube to follow non-linear / tracking camera moves.
06activitymotion per slice
little motion
lots of motion
Click the strip to jump the cube's time cursor.
07volume stats
dimensions
—
slices
—
strategy
—
mean motion
—
▶playback2D source
t = 0.00s
dur 0.00s
nearest slice—
selected region
—
—
?how it works
A video becomes a spatial-temporal voxel volume — every voxel carries RGB color and an opacity α.
• dynamic voxels → high α (glow)
• static voxels → low α (ghost)
• boundary voxels → face outlines
• alpha voxels → pre-keyed footage keeps its own transparency
Interactions: rotate to remove self-occlusion · shear for panning synopses · deform to follow tracking shots · click the volume to slice straight to that moment.