Ljómi uses the aurora’s spectral characteristics to map brightness regions to musical scales and wave patterns to CC parameters, generating real-time MIDI output with configurable HSV thresholds and wave coherence analysis.

System Architecture

Ljómi uses advanced computer vision technology to detect and analyze aurora characteristics, translating these into musical elements through a sophisticated processing pipeline:

Video Processing

The system begins by capturing aurora footage through either a camera (for live performances) or pre-recorded video files. The Video Processor serves as the central orchestrator, efficiently managing frame acquisition and coordinating all analysis components.

Colour & Pattern Analysis

Ljómi identifies the three primary aurora colours - green, pink, and purple - analysing their:

  • Intensity: The brightness and prominence of each colour

  • Movement: Direction and speed of motion across the frame

  • Wave patterns: Coherence, wavelength, and organisation of aurora movements

  • Area coverage: Percentage of the frame covered by each colour

Musical Translation

These visual characteristics are transformed into musical elements through:

  • Note generation: Based on brightness regions within each colour

  • Chord complexity: Evolving from simple triads (low brightness) to rich ninth chords (high brightness)

  • Control signals: Wave direction, speed, coherence, and other parameters mapped to MIDI CC values

  • Musical scales: Extensive selection from traditional to world music scales

Key Configurable Parameters

Ljómi offers extensive customisation through its configuration system:

  • Video Settings

    • Input source: Camera or video file

    • Processing resolution: From 240p (very fast) to HD (high detail)

    • Frame skip rate: Process every Nth frame for performance optimisation

    • Display options: Configurable visualisation overlays

  • Color Detection

    • HSV ranges: Fine-tune detection for each aurora colour

    • Brightness thresholds: Minimum brightness for detection

    • Area requirements: Minimum size for meaningful detection

  • Musical Configuration

    • Scale selection: From traditional (major/minor) to exotic (Hungarian minor, Japanese, Persian)

    • Root note and octave: Base musical foundation

    • Note generation mode: Single notes or chord progressions

    • MIDI channel assignment: Per-colour channel routing

  • Wave Analysis

    • Sensitivity settings: Control response to subtle movements

    • Coherence thresholds: Detect level of wave organisation

    • Direction stability: Smooth transitions between movement patterns

Performance Optimization

Ljómi employs several strategies to ensure smooth operation:

  • Intelligent Processing

    • Resolution scaling: Separate processing and display resolutions

    • Frame skipping: Process only key frames for natural musical transitions

    • Value deduplication: Prevent MIDI message flooding