Audiobook Generator

This project is an automated pipeline for scraping, processing, and converting web novel chapters into high-quality audiobooks. It supports multiple sources and advanced audio processing, including effects and playback speed adjustment.

---

Features

• Scrape chapters from RoyalRoad and ScribbleHub
• Generate speech using Qwen3 TTS (default) or Coqui TTS
• Merge and apply audio effects with ffmpeg
• Change playback speed
• Designed for batch processing and meta-progression tracking
• Easily configurable via config.yml

---

Installation

1. Clone the repository

git clone https://github.com/robert-clayton/audiobook.git
cd audiobook

2. Install uv (if not already installed)

Windows:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Linux/macOS:

curl -LsSf https://astral.sh/uv/install.sh | sh

3. Install dependencies

uv sync

---

Usage

uv run audiobook [--speed 1.2] [--dev]

Options

• --speed: Playback speed multiplier (default: 1.0)
• --dev: Use the development config file (config_dev.yml)

---

Configuration

Edit config.yml or config_dev.yml:

config:
  output_dir: \\192.168.0.1\your\mapped\network\location\media
series:
  - latest: https://www.royalroad.com/fiction/36049/the-primal-hunter/chapter/2309548/chapter-1095-beauty-brain
    name: The Primal Hunter
    narrator: onyx
    url: https://www.royalroad.com/fiction/36049/the-primal-hunter
  - latest: https://www.royalroad.com/fiction/24709/defiance-of-the-fall/chapter/2307815/chapter-1329-past-lifes-dream
    enabled: false
    name: Defiance of the Fall
    narrator: onyx
    url: https://www.royalroad.com/fiction/24709/defiance-of-the-fall
  - latest: https://www.royalroad.com/fiction/76463/mage-tank/chapter/2306656/273-the-truth-revealed
    name: Mage Tank
    narrator: onyx
    replacements:
      DR: Damage Reduction
    system:
      speed: 1.1
      modulate: true
      type:
      - bold
      voice: fable
    url: https://www.royalroad.com/fiction/76463/mage-tank

Each series entry includes:

• name: Display name
• url: Table of Contents url
• latest: Latest chapter url
• narrator: What voice to use for TTS synthesis. Must match name (sans ext) of files in speakers/
• enabled: Toggle scraping for this series (optional, default True)
• replacements: Optional strings to replace from the source material (optional)
• system: Identify certain divs as from a "system" to modulate the audio for (optional)
  • Currently supported types: bold, italic, bracket, braces, table, center, angle, blockquote
  • Modulate adds a "tech-y" filter to the voice
  • Speed will speed up and down the system voice

---

Structure

audiobook/
├── scrapers/         # RoyalRoad and ScribbleHub specific scrapers
├── processors/       # TTS, audio merging, playback speed adjustments
├── utils/            # Logging, color codes, audio helpers
└── __main__.py       # CLI entry point
speakers/             # User-provided source audio for voice cloning
config.yml            # User-editable config

---

Troubleshooting

FFmpeg Errors

Ensure ffmpeg is installed and accessible via your system PATH. You can verify with:

ffmpeg -version

---

License

MIT License. See LICENSE file for details.