Speech Analysis Module (SAM)

Speech Analysis Module is a modular Python library designed to simplify audio preprocessing, feature extraction, visualization, and recording in experimental and clinical psychology settings.

This package provides ready-to-use functions for:

Recording audio

• Compatible with USB microphones and Arduino triggers
• Includes classes for simple and Arduino-synced recordings

Preprocessing audio

• Bandpass filtering and noise gating
• Silence trimming and duration standardization

Analysis

• Onset detection
• Power spectral density (PSD) via STFT or Welch’s method
• Envelopes, frequency analysis, and energy-based features

Visualization

• Oscillograms
• Spectrograms
• PSD curves in both dB and linear scale

Features
-Integrated analysis with Praat via Parselmouth: Automatic extraction of pitch, jitter, shimmer, harmonicity (HNR), and other acoustic biomarkers from speech recordings -Acoustic and spectral feature extraction with librosa: Includes MFCCs, spectral centroid, spectral bandwidth, spectral contrast, spectral rolloff, zero-crossing rate, RMS energy, chroma features, and tonal centroid features (tonnetz)

Structure

speech_analysis_module/
│
├── analysis/       # Core signal processing (onsets, PSD, Welch)
├── plot/           # Visualization functions (waveforms, spectrograms)
├── prepro/         # Preprocessing utilities (filters, trimming, noise gate)
├── utils/          # File I/O, transcription (Whisper), helper functions
├── psychopy/       # Recording classes for PsychoPy experiments
└── features/       # (Coming soon) Feature extraction for voice analysis


## Installation

You can install SAM directly from PyPI:

```bash
pip install speech-analysis-module