Side-by-side comparisons of the Filian avatar rendered at 8, 12, 24, 30, and 60 fps — evaluating how downsampled frame rates affect perceived motion, readability, and stylization across walking, running, and acrobatic clips.
This research explores how rendering the Filian avatar at reduced frame rates changes its visual character. Each clip is a five-panel comparison showing the same motion simultaneously at 8, 12, 24, 30, and 60 fps, letting us directly compare how each rate handles different types of movement.
The frame rate cap is applied at OBS, not anywhere earlier in the pipeline. Warudo continues to render at its native internal rate — OBS is the sole place where the output frame rate is constrained, captured, and encoded.
This means nothing upstream of OBS (Warudo, the mocap pipeline, NiloToon) is being altered. All visual differences below are a product of OBS-side capture frame rate.
| Property | Value |
|---|---|
| Avatar | Filian 1.4.0 |
| Render Pipeline | URP |
| Scene File | Filian 1.4.0 (URP).warudo |
| Motion Capture | OptiTrack |
| FPS Cap Location | OBS (not Warudo / not mocap) |
| Compared Rates | 8 / 12 / 24 / 30 / 60 fps |
| Date Captured | 2026-04-15 |
| Total Clips | 17 |
Continuous, low-velocity locomotion. Lower frame rates tend to read well here — the pose-to-pose rhythm of walking forgives stepped playback.
High-velocity locomotion with faster limb travel. Low frame rates exaggerate stepping and can introduce visible temporal gaps during fast limb motion and abrupt direction changes.
Fast, full-body rotations and inversions. These are the hardest clips for low frame rates to handle — motion blur is absent, so any gap between captured frames reads as a clean teleport.
Isolating secondary motion (hair sim) from body motion. Low frame rates affect high-frequency simulated motion very differently than deliberate keyframed body animation.