Video Compressor

Video Compressor uses FishID's FFMPEG preferred command to reduce file size and optimize it for analysis workflows. This project provides both direct FFmpeg usage and a Docker container for easy deployment and sharing.

FFmpeg Command

ffmpeg -hide_banner -loglevel error -i INPUT.MP4 -r 24 -vf scale=1920:-2,setsar=1:1 -c:v libx264 -crf 25 -preset faster -an -movflags +faststart INPUT-TRANSCODED.mp4

This command does:

• Changes frames rate to 24 FPS, assumes that frame rate is over 24 FPS, -r 24
• Resize resolution to 1920 width keeping aspect ratio, -vf scale=1920:-2,setsar=1:1
• Sets video codec to H.264 using the libx264 encoder, which is widely compatible and offers good compression -c:v libx264
• Sets image quality, Constant Rate Factor (CRF), a quality setting used with libx264 to 25, -crf 25, crf ranges from 0 which is equivalent to lossless, 18–23 which visually is a very good quality, to 51 very bad quality
• Controls encoding speed vs compression efficiency, -preset faster
• Removes audio track, -an
• Moves the MP4 metadata (moov atom) to the beginning of the file, which enables fast streaming / playback in browsers and on the web.

Docker Usage

Using Pre-built Images (Recommended)

The easiest way to use this tool is with our pre-built Docker images:

Option 1: GitHub Container Registry (Free)

docker pull ghcr.io/your-username/video-compressor:latest
docker run --rm \
  -v /path/to/your/videos:/app/input:ro \
  -v $(pwd)/output:/app/output \
  ghcr.io/your-username/video-compressor:latest

Option 2: Docker Hub

docker pull your-username/video-compressor:latest
docker run --rm \
  -v /path/to/your/videos:/app/input:ro \
  -v $(pwd)/output:/app/output \
  your-username/video-compressor:latest

Building Locally

If you prefer to build the image yourself:

1. Build the Docker image:

   docker build -t video-compressor .

2. Process all MP4 files in a directory (Batch Mode):

   # Create output directory
   mkdir -p output
   
   # Run batch processing
   docker run --rm \
     -v /path/to/your/videos:/app/input:ro \
     -v $(pwd)/output:/app/output \
     video-compressor

3. Process a single file:

   docker run --rm \
     -v /path/to/your/videos:/app/videos \
     video-compressor \
     /app/videos/input.mp4 /app/videos/output-transcoded.mp4

Using Docker Compose

1. Setup:

   # Create directories
   mkdir -p videos output
   
   # Copy your MP4 files to the videos directory
   cp /path/to/your/*.mp4 videos/

2. Run batch processing:

   docker-compose up video-compressor

3. Run single file processing:

   docker-compose --profile single run video-compressor-single input.mp4 output.mp4

Volume Mounting Options

• Batch processing: Mount your video directory to /app/input (read-only) and output directory to /app/output
• Single file processing: Mount your video directory to any path inside the container and specify full paths

Examples

Batch process all videos in /Users/john/videos:

docker run --rm \
  -v /Users/john/videos:/app/input:ro \
  -v $(pwd)/output:/app/output \
  video-compressor

Process specific file with custom output name:

docker run --rm \
  -v /Users/john/videos:/app/videos \
  video-compressor \
  /app/videos/my-video.mp4 /app/videos/compressed-video.mp4

Features

• Batch Processing: Automatically processes all MP4 files in a directory
• Single File Processing: Process individual files with custom output names
• Docker Support: Easy deployment and sharing with collaborators
• Automated Builds: GitHub Actions automatically builds and publishes Docker images
• Multi-platform: Supports both Intel/AMD (x64) and ARM64 (Apple Silicon) architectures
• Cross-platform: Works on any system with Docker installed
• No Dependencies: All required tools (FFmpeg) are included in the Docker image

Automated Builds

This project uses GitHub Actions to automatically build and publish Docker images:

• 🚀 Automatic: Images are built on every push to the main branch
• 🏷️ Versioned: Tagged releases create versioned Docker images
• 🔄 Multi-platform: Built for both x64 and ARM64 architectures
• 📦 Multiple registries: Published to both GitHub Container Registry and Docker Hub

See SETUP.md for detailed instructions on configuring automated builds.