Skip to main content
All experiments
Audio Visual
Live

Chladni Visualizer

Audio-reactive standing-wave renderer that maps waveform features into Chladni modes, palette shifts, beat flashes, and exportable social clips.

The visualizer turns audio into a simulated resonant plate. A librosa analyzer extracts waveform, FFT, tempo, onset, chroma, MFCC, and band-energy features. Those features drive mode selection, intensity, scale, sharpness, palette warmth, and beat-responsive flashes.

The physics layer renders standing-wave superpositions with numba acceleration, then a color engine maps the signed pattern into HSL palettes. The CLI can export MP4, WebM, GIF, PNG previews, and title overlays; the browser lab version gives the same idea as an immediate drag-and-drop preview.

This sits best as a lab artifact first because the useful object is interactive: drop a track, watch the plate respond, tune a preset, export a frame or clip. The field note records the implementation choices, the false starts, and why nearest-neighbor resizing preserved the nodal lines better than a smoother filter.

Highlights

  • Web Audio preview with drag-and-drop input, presets, typography overlay, PNG export, and WebM capture
  • Python render pipeline: librosa analysis, feature mapping, numba Chladni simulation, HSL color mapping
  • Presets for classic plate studies, festival visuals, album art, and social clips
  • Field note paired with the live artifact so the implementation is inspectable, not just decorative

Stack

  • Python
  • librosa
  • numba
  • Canvas
  • Web Audio
  • FFmpeg