← Back to Portfolio
JUCE · C++ · Psychoacoustic Analysis In Development

Acoustic
Analyzer

A perception-first analysis plugin that goes beyond raw frequency readouts. Calculates an Acoustic Activation Index (0–100) representing how stimulating or calming a signal is likely to feel to a listener — combining brightness, harshness, dynamic variability, and temporal unpredictability into a single interpretable score with CSV export for research use.

🎛️ VST3 🍎 AU 🖥️ Standalone 64-bit 💾 CSV Export
01 — Demo

Plugin Demo

// replace with YouTube embed when ready

02 — Interface
🎛️

Main interface — colour-coded Activation Score + metric bars

📈

Real-time metric graphs + historic traces

🔴

Recording session & CSV export

🧪

Standalone — research & quick testing

📊 Analysis Metrics

  • Spectral Centroid (Brightness) — perceptual centre of gravity, normalised 0–8 kHz
  • Spectral Harshness — high-frequency energy ratio (>2 kHz) scaled to perceptual impact
  • Dynamic Variability — RMS variance over a moving window capturing loudness unpredictability
  • Temporal Unpredictability — successive RMS-change statistics detecting sudden transients

💾 Research & Export

01

Real-time recording

Start/stop with live timestamped frames at ~23–30 Hz. Per-frame metrics captured in a thread-safe buffer, ready for export at any point.

02

CSV export

Session exports include precise timestamps, the composite activation score, and all component metrics. Format ready for R, Python, or Excel pipelines.

03

Research pairing

Designed to correlate acoustic activation scores with questionnaire responses or physiological measurements in controlled listening studies.

04 — Acoustic Activation Index

Four perceptual dimensions combine into a single 0–100 score. Green = calm, red = activating. Coefficients derived from research into arousal responses to spectral and temporal audio features.

Harshness 35%

Strongest single predictor — high-frequency harshness triggers arousal responses most reliably.

Brightness 25%

Spectral centroid position — brighter sounds score higher on alertness and stimulation scales.

Dynamic Variability 20%

Loudness unpredictability over time — higher variance keeps the auditory system in an alert state.

Temporal Unpredictability 20%

Sudden transient detection — startle-pattern triggers contribute independently of overall loudness.

05 — Download

Beta builds posted during testing phases. Contact for early access or research partnerships.

🪟
VST3 — Windows
64-bit Windows build
Coming soon
🍎
VST3 — macOS
Universal (Intel + Apple Silicon)
Coming soon
🎵
AU — macOS
Logic / GarageBand
Coming soon
🖥️
Standalone
No DAW required
Coming soon
06 — Technical Specs
C++ / JUCE
Core engine
2048-pt Hann FFT
Frequency analysis
64-bit floats
Thread-safe processing
44.1–192 kHz
Supported sample rates
30 Hz UI refresh
Recommended for smooth visuals
CSV export
Timestamped per-frame session data

🔬 Planned Enhancements

A-weighting, ISO-compliant sharpness/roughness metrics, Bark/ERB conversions, SPL calibration, and expanded visualisation modes for publication-ready output figures.

Interested in Collaboration?

Open to beta testers, data-sharing arrangements, and research partnerships involving acoustic perception measurement.