Audio Enhancement Workflow

This project provides a complete workflow for enhancing audio files using resemble-enhance, with support for various input and output formats.

Overview

The workflow consists of four main steps:

1. Convert input audio file(s) to WAV format
2. Enhance the audio using resemble-enhance
3. Convert the enhanced audio to the desired output format
4. Save (together with original metadata) the enhanced audio to destination directory

Usage

./run.sh <source_dir> <target_dir> [format] [quality]

Parameters

• source_dir (required): Directory containing input audio files
• target_dir (required): Directory for final processed audio files
• format (optional): Output format (defaults to flac)
• quality (optional): Audio quality for lossy formats (defaults to 256k)

Examples

# Process with default settings (FLAC output)
./run.sh input_dir output_dir

# Process to MP3 with specific quality
./run.sh input_dir output_dir mp3 320k

# Process to WAV
./run.sh input_dir output_dir wav

# Process to Opus
./run.sh input_dir output_dir opus 128k

Supported Formats

Input Formats

• MP3, M4A, AAC, FLAC, OGG, WMA, AIFF, AU, RA, 3GP, AMR, OPUS, WAV

Output Formats

• FLAC (default, lossless)
• MP3 (with libmp3lame)
• AAC
• OGG (with libvorbis)
• M4A (AAC in MP4 container)
• WAV (PCM 16-bit 44.1kHz)
• OPUS (with libopus)

Performance

Processing time depends on your hardware. As a reference:

• GPU: GeForce RTX 3060 Ti: Approximately 1 hour of processing time per hour of audio
• CPU: 2,3 GHz I’m 8-Core intel Core i9: Approximately 11 hours of processing time per hour of audio

Requirements

• Linux operating system or wsl (this script uses bash and Linux-specific commands)
• Python 3.10
• ffmpeg (for audio conversion)
• resemble-enhance (for audio enhancement)

Installation

1. Install ffmpeg:

   # Ubuntu/Debian
   sudo apt update && sudo apt install ffmpeg
   
   # Other Linux distributions
   # Use your distribution's package manager (yum, pacman, zypper, etc.)

2. Create and activate a Python virtual environment:

   python3.10 -m venv .venv
   source .venv/bin/activate

3. Install resemble-enhance:

   pip install resemble-enhance

4. Make the script executable:

   chmod +x run.sh

Script Details

Main Scripts

• run.sh: Main orchestration script
• scripts/convert_audio_to_wav.py: Converts various audio formats to WAV
• scripts/convert_wav_to_format.py: Converts WAV files to target format

Workflow Steps

1. Audio Conversion: All input files are converted to WAV format (16-bit, 44.1kHz)
2. Enhancement: resemble-enhance processes the WAV files for noise reduction and quality improvement
3. Format Conversion: Enhanced WAV files are converted to the specified output format and quality
4. Cleanup: Temporary files are cleaned up (with user confirmation for enhanced WAV files)

Quality Settings

Quality settings are only applicable for lossy formats:

• MP3: 128k, 192k, 256k, 320k
• AAC: 128k, 192k, 256k, 320k
• OGG: 128k, 192k, 256k, 320k
• Opus: 64k, 96k, 128k, 192k

Lossless formats (FLAC, WAV) do not use quality settings.

Troubleshooting

Common Issues

1. ffmpeg not found: Install ffmpeg using your system's package manager
2. resemble-enhance not found: Install using pip install resemble-enhance
3. Permission denied: Make sure the script is executable with chmod +x run.sh

Error Messages

The script provides detailed error messages for common issues:

• Missing source directory
• Failed audio conversion
• Enhancement process failures
• Output format conversion issues