ffmpeg-batcher

A lightweight, queue-based automation framework for batch-converting videos and images using FFmpeg.

It is designed to handle heavy media conversion tasks sequentially, preventing system overload while maintaining a clean, tracked workflow.

---

Features

• Queue-Based Workflow: Process heavy video/image conversions one by one without overloading your system.
• Ready-to-Use Presets: Includes handy presets for AV1 (libaom, libsvtav1), WebP, Opus, and custom video trimming.
• Robust Processing (run.sh): * Automatically adjusts process priority (renice) to keep your OS responsive during intense CPU encodes.
  • Saves temporary outputs to a work-in-progress (wip) directory, moving them to fin only upon successful completion.
  • Safely removes completed queue items in real-time.
  • Logs the entire process to log.txt with timestamps.

---

Project Structure

.
├── original/               # Place your raw, unconverted files here
├── queue/                  # Symlinks to original files, representing the active queue
├── wip/                    # Temporary directory for files currently being processed
├── fin/                    # Final directory where successfully converted files are stored
├── select.sh               # Selects the active FFmpeg preset script
├── gen_queue.sh            # Generates symlinks from original/ to queue/
├── run.sh                  # The main batch-processing engine
├── print_info.sh           # Helper script to output video frame counts
├── video_*.sh / image_*.sh # Pre-configured FFmpeg scripts
└── log.txt                 # Main execution logs

---

Prerequisites

Ensure you have ffmpeg, ffprobe, and bash installed on your system.

• Arch Linux: sudo pacman -S ffmpeg bash
• Debian/Ubuntu: sudo apt install ffmpeg bash
• macOS (via Homebrew): brew install ffmpeg

---

Quick Start

Here is a quick guide on how to batch-convert multiple images to the modern .webp format:

1. Set Up Your Files

Create the original directory if it doesn't exist, and drop your input files inside:

mkdir -p original
# Put your source files (e.g., .png, .jpg) inside original/

2. Select a Preset

Run select.sh with your desired preset script. This creates a symlink named function pointing to the active preset:

./select.sh image_libwebp.sh

3. Generate the Queue

Generate symlinks from your original/ folder to the queue/ folder to build the pipeline:

./gen_queue.sh

4. Run the Batch Process

Start the conversion engine:

./run.sh

Note: The script may prompt you for your sudo password to lower its CPU priority (renice). This ensures your computer remains perfectly usable while FFmpeg runs in the background.

---

How It Works Under the Hood

The conversion pipeline operates in a safe, step-by-step cycle:

[original/] ---> (gen_queue.sh) ---> [queue/] ---> (run.sh + preset) ---> [wip/] ---> [fin/]

1. Queue Generation (gen_queue.sh): Instead of copying heavy media files and wasting disk space, it creates symbolic links inside queue/ pointing back to your original/ files.
2. Preset Selection (select.sh): It symlinks your chosen conversion script to a file named function. This allows run.sh to trigger the active script dynamically using ./function without hardcoding the format.
3. Process Optimization (run.sh): When you start run.sh, it uses sudo renice to assign a lower priority value to the current process. This is incredibly helpful for long, resource-heavy encodes (like AV1), keeping your UI snappy.
4. Safe Processing Cycle: For every file in the queue:

• run.sh logs the metadata using ./print_info.sh.
• The preset is executed, and the intermediate file is written to wip/.
• Upon successful completion (exit status 0), the original queue link is safely deleted, and the finalized file is moved to fin/.
• If a crash or interruption occurs, your files remain safely in wip/ or queue/ so you never lose progress.

---

License

This project is licensed under the GPL-3.0 License. See the LICENSE file for details.