RhythmEngine 🎵

A Python library for audio rhythm analysis and click track generation. Detect beats in audio files and generate synchronized click tracks with dynamic volume levels.

Features

• Beat Detection: Multi-processed analysis for accurate beat tracking
• Click Track Generation: Create dynamic click sounds that sync with detected beats
• Audio Export: Save processed audio with click tracks in standard formats
• Customizable Parameters: Adjust click sounds and analysis parameters

Installation

Requirements

• Python 3.8+
• FFmpeg (for audio file support)

Dependencies

pip install librosa numpy pydub

Recommended Installation

# Clone repository
git clone https://github.com/andriy8800555355/RhythmEngine.git
cd RhythmEngine

# Install requirements
pip install -r requirements.txt

Usage

Basic Example

from RhythmEngine import RhythmEngine

# Initialize with audio file
engine = RhythmEngine("input_song.mp3")

# Generate click track
mixed_audio = engine.generate_click_track()

# Save output
engine.save_mixed_audio("output_with_clicks.wav")

# Get detected beats
beat_times = engine.get_beat_times()
print(f"Detected beats at: {beat_times}")

Advanced Usage

# Custom click sound parameters
engine.generate_click_track(
    click_duration=0.1,  # Click duration in seconds
    freq1=1000,          # Primary frequency (Hz)
    freq2=1500           # Secondary frequency (Hz)
)

# Access raw audio data
original_audio = engine.y          # Raw audio waveform
sample_rate = engine.sr            # Sampling rate
mixed_audio = engine.mixed_audio   # Processed audio with clicks

API Documentation

RhythmEngine(audio_path: str)

Main class for rhythm processing

Parameters:

• audio_path: Path to audio file (supports most common formats)

Methods

generate_click_track(click_duration=0.07, freq1=880, freq2=1320)

Generate click track synchronized with detected beats

Parameters:

• click_duration: Duration of click sound in seconds (default: 0.07)
• freq1: Primary frequency component in Hz (default: 880)
• freq2: Secondary frequency component in Hz (default: 1320)

Returns:

• np.ndarray: Mixed audio waveform with click track

save_mixed_audio(output_path: str)

Save processed audio to file

Parameters:

• output_path: Output file path (format determined by extension)

get_beat_times()

Get detected beat times

Returns:

• List[float]: List of beat times in seconds

Implementation Details

• Multi-processing: Uses parallel processing for faster beat detection
• Dynamic Volume: Click volume adapts to onset strength of original audio
• Normalization: Automatic peak normalization prevents clipping
• Tempo Estimation: Combines global tempo estimation with local beat tracking

Supported Formats

Input: MP3, WAV, FLAC, OGG, AAC (any format supported by FFmpeg)
Output: WAV (other formats supported via Pydub)

Acknowledgments

• Uses librosa for audio analysis
• Audio processing with pydub
• Built with NumPy for efficient array operations

---

Contributions welcome! Please open an issue or PR for suggestions/bug reports.