All credit goes to Colugo - Here's his post showcasing the script

sonic-garbage

Overview

Sonic Garbage is a Python script that automates the process of generating audio samples from YouTube videos. It randomly selects search phrases from a provided word list, downloads corresponding YouTube videos, and processes the audio to create one-shot and looped versions. The processed audio files are then saved in designated directories.

Features

• Random Search Phrases: The script generates random search phrases by selecting two words from a provided word list.
• YouTube Video Download: Utilizes the yt_dlp library to download YouTube videos based on the generated search phrases.
• Audio Processing: Creates both one-shot and looped versions of the downloaded audio, applying fade-in, fade-out, and normalization.
• Batch Processing: Executes the entire process in batches, allowing for the creation of multiple audio samples in one run.

Prerequisites

• Python 3
• Required Python libraries (pydub, yt_dlp, isodate)
• FFmpeg (See FFmpeg Installation)

Usage

1. Clone the repository to your local machine.
2. Install the required dependencies using pip install -r requirements.txt.
3. Run the script using python sonic-garbage.py.
4. Check the wavs/processed/loop and wavs/processed/oneshot directories for the processed audio files.

Disclaimer

Please ensure that your use of this script complies with YouTube's terms of service. Downloading YouTube videos may be subject to legal and ethical considerations.