PylonGUY

Your best friend for high-performance camera control for Basler cameras with high-speed recording capabilities and a DropletUI frontend.

Features

• Real-time preview with adaptive grab strategy (lag-free at any FPS)
• Camera configuration with automatic adjustment to camera capabilities
• High-speed frame capture to raw frames with FFmpeg post-processing
• Configurable playback rate for slow-motion video
• Interactive ROI selection and offset controls
• Preset configurations (Default, HighSpeed, FullFrame, Microfluidics)
• Auto-apply presets on camera connection
• Waterfall mode for line-scan recording

Installation

From PyPI

pip install pylonguy

From source

git clone https://github.com/merv1n34k/pylonguy.git
cd pylonguy
pip install -e .

Requirements

• Python 3.12+
• ffmpeg in PATH (for video processing)
• Basler Pylon SDK (for camera communication)

Installing ffmpeg

# macOS
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

# Windows
# Download from https://ffmpeg.org/download.html and add to PATH

Installing Pylon SDK

Download from Basler website. See pypylon known issues for platform-specific notes.

Usage

# Run the application
pylonguy

# Or with uv
uv run pylonguy

Presets

Camera presets are saved to presets.json in the current working directory. This allows you to maintain different presets for different projects by running the application from each project's folder.

How it works

The application separates capture from video encoding for maximum performance:

1. Capture Phase: Frames stream from camera → memory buffer → raw files on disk
2. Processing Phase: After recording stops, FFmpeg assembles raw frames into video

This allows you to specify any desired playback frame rate, e.g.:

• Camera captures at 1000 fps
• Video plays back at 24 fps
• Result: 1 second of capture = 41.7 seconds of slow-motion video

Waterfall Mode

Record single lines instead of full frames for:

• Maximum frame rates (minimal data transfer)
• Line-scan camera emulation
• Monitoring flow in microfluidic channels

Output format: .wtf (waterfall) files, convertible to PNG:

python -m pylonguy.wtf2png recording.wtf

License

Distributed under the MIT License. See LICENSE for more information.