Create Content Aware Scale videos in minutes, not hours.

A high-performance CLI tool for Content Aware Scaling (a.k.a liquid rescaling / seam carving) of images and videos. Built to make Content Aware Scale meme videos actually practical to create. What used to take hours in Photoshop now takes minutes with parallel processing and native video support.

Click to watch demo video

Ever seen those stupid Content Aware Scale videos on YouTube before? I'm responsible for a few of those... this one in particular: Content Aware Scale - "Show me your war face!". The problem? Each video took hours to create.

While Neil Cicierega's Animated Content Aware Scaling script for Adobe Photoshop helped automate the process, it was still painfully slow since Photoshop only processes one frame at a time. For longer videos, this meant waiting hours for Photoshop to finish processing each frame sequentially.

I built cascaler to solve this bottleneck. What used to take hours of waiting for Photoshop now completes in minutes with parallel batch processing.

• Parallel processing - Process multiple frames simultaneously
• Native video support - No more exporting/importing image sequences manually
• Significantly faster than the Photoshop script workflow

• Gradual scaling - Smooth* transitions from 100% to any percentage over time
• Audio effects - Built-in vibrato and tremolo filters for extra weirdness (extra customization to come)
• Frame-accurate trimming - Process specific segments without re-encoding the entire video
• Scale-back mode - Apply the liquid rescaling effect then scale back to original dimensions
• Audio preservation - AAC audio encoding with proper sync (extra codecs coming soon)
• Progress tracking - Real-time progress bars with ETA
• Persistent configuration - Save your preferred settings

_{*As smooth as seam carving can be...}

• .NET 10.0 or higher
• FFmpeg 7.x libraries (for video processing)

# Clone and build
git clone https://github.com/nathanpbutler/cascaler.git
cd cascaler
dotnet build

# Install FFmpeg (if not already installed)
brew install ffmpeg@7              # macOS
sudo apt install ffmpeg            # Linux
# Windows: Download from https://www.gyan.dev/ffmpeg/builds

For detailed FFmpeg setup instructions, see Configuration Guide.

# Scale a single image to 50% (default)
cascaler input.jpg

# Scale to specific dimensions
cascaler input.jpg -w 800 -h 600

# Process an entire directory
cascaler /path/to/images -p 75

# Process a video (preserves audio)
cascaler input.mp4 -o output.mp4 -p 75

# Trim and process a video segment
cascaler input.mp4 -o output.mp4 --start 10 --duration 5 -p 50

# Add audio effects
cascaler input.mp4 -o output.mp4 --vibrato -p 75

# Gradual scaling effect (100% → 50%)
cascaler input.mp4 -o output.mp4 -sp 100 -p 50

# Image to video with gradual scaling
cascaler input.jpg -o output.mp4 --duration 3 -sp 100 -p 50

# Apply effect and scale back to original size
cascaler input.jpg --scale-back -p 50

# Convert directory to video with gradual scaling
cascaler /path/to/images -o output.mp4 -sp 75 -p 25

For complete command-line reference and advanced usage, see Command-Line Reference.

For a complete list of options, see Command-Line Reference.

Save your preferred settings to avoid repeating command-line options:

# Initialize configuration with FFmpeg detection
cascaler config init --detect-ffmpeg

# View current configuration
cascaler config show

# Show config file location
cascaler config path

Configuration file location:

• Unix/macOS/Linux: ~/.config/cascaler/appsettings.json
• Windows: %APPDATA%\cascaler\appsettings.json

Example configuration:

{
  "Processing": {
    "MaxImageThreads": 32,
    "DefaultScalePercent": 75,
    "DefaultFps": 30
  },
  "VideoEncoding": {
    "DefaultCRF": 20
  },
  "Output": {
    "Suffix": "-scaled"
  }
}

Settings priority: Embedded defaults → User config → Command-line arguments

For detailed configuration options and examples, see Configuration Guide.

Input:

• Images: .jpg, .png, .gif, .bmp, .tiff, .webp, .ico
• Videos: .mp4, .avi, .mov, .mkv, .webm, .wmv, .flv, .m4v

Output:

• Videos: .mp4, .mkv (H.264 encoding with AAC audio)
• Frames: png (default), jpg, bmp, tiff

Built with modern .NET 10.0 and powered by:

• ImageMagick.NET for seam carving
• FFmpeg.AutoGen 7.1.1 for native video/audio processing
• Parallel processing (configurable: 16 threads for images, 8 for video)

See Architecture Documentation for technical details.

• Command-Line Reference - Complete guide to all command-line options and usage patterns
• Configuration Guide - Persistent configuration setup, FFmpeg path detection, and advanced settings
• Architecture Documentation - Technical overview, processing flows, and implementation details

Contributions are welcome! Please feel free to submit issues, feature requests, or pull requests.

MIT License. See LICENSE for details.