🎬 ASMR Pro Cutter

Automatic tool to cut and create ASMR clips from long videos, with AI-based intelligent audio analysis to identify the best moments.

✨ Features

• 🎯 Intelligent audio analysis - Automatically detects the best ASMR triggers (clicks, taps, crunches)
• 🎨 Modern graphical interface - Intuitive GUI with tkinter
• ⚡ GPU acceleration - NVIDIA NVENC support for ultra-fast encoding
• 🎞️ Quality preserved - Maintains source video quality (2K/4K)
• 📊 Customizable parameters - Adjustable clip duration, pre/post-roll
• 🗂️ Organized output - Automatically saves clips with ordered timestamps

🖥️ Interface

🎵 How it works

The program analyzes video audio using three metrics:

1. Spectral Centroid - Identifies sound brightness
2. Onset Strength - Detects trigger impact and suddenness
3. Zero Crossing Rate - Finds sharp metallic/plastic sounds

It combines these parameters to create a "crispness" index and automatically selects the best moments.

📺 Example Result

See what ASMR Pro Cutter can do! This Short was automatically generated from a 5-minute video:

Watch on YouTube

📦 Installation

Prerequisites

• Python 3.8 or higher
• NVIDIA GPU (optional, for hardware acceleration)
• FFmpeg (automatically installed with imageio-ffmpeg)

Setup

1. Clone the repository

git clone https://github.com/lordpba/ASMR-Pro-Cutter.git
cd ASMR-Pro-Cutter

2. Create a virtual environment (recommended)

python -m venv .venv

3. Activate the virtual environment

Windows:

.venv\Scripts\activate

Linux/Mac:

source .venv/bin/activate

4. Install dependencies

pip install .

Or for development:

pip install -e .

🚀 Usage

Graphical Interface (Recommended)

python gui.py

1. Click "Select Video" and choose your ASMR video
2. (Optional) Select a custom output folder
3. Adjust parameters if needed:
   • Total target duration: Total duration of combined clips (~58s for Shorts)
   • Pre-roll: Seconds before trigger (default 1.2s)
   • Post-roll: Seconds after trigger (default 1.3s)
   • Final clip extra: Extra seconds for last clip closing shot (default 2.0s)
4. Click "START PROCESSING"
5. Clips will be saved to videoname_shorts/ with ordered names

Command Line

python main.py

Place videos in video_input/ and the program will automatically process all found video files.

📁 Output Structure

your_video_shorts/
├── clip_001_at_0045s.mp4
├── clip_002_at_0123s.mp4
├── clip_003_at_0189s.mp4
└── ...

Files are named with progressive number and timestamp for easy sorting in video editors.

⚙️ Advanced Parameters

You can modify directly in main.py or use the GUI's advanced section:

TARGET_DURATION = 58.0   # Total target duration (seconds)
PRE_ROLL = 1.2          # Seconds before trigger
POST_ROLL = 1.3         # Seconds after trigger
FINAL_CLIP_EXTRA = 2.0  # Extra seconds for last clip
MIN_FREQ = 1800         # Minimum frequency for filtering (Hz)
HOP_LENGTH = 512        # Audio analysis precision

GPU/Quality Parameters

In encoding code (line ~146):

• CQ Value (-cq 18): Constant quality

  • 0 = Lossless (huge files)
  • 18 = Near lossless, excellent compromise (default)
  • 23 = High quality
  • 28 = Medium quality

• NVENC Preset (preset="slow"):

  • fast = Fast, medium quality
  • medium = Balanced
  • slow = Maximum quality (default)

🎥 Supported Formats

• Input: MP4, MOV, AVI, MKV
• Output: MP4 (H.264 + AAC)
• Audio: AAC 320kbps (maximum quality for ASMR)
• Video: H.264 NVENC CQ18 (near lossless)

💡 Tips

• For 4K videos, ensure you have at least 8GB of RAM
• GPU encoding is ~10-20x faster but requires NVIDIA GPU
• Use longer pre-roll (2-3s) for videos with slow movements
• Reduce TARGET_DURATION for more selective clips

🐛 Troubleshooting

Error "No module named 'moviepy'"

pip install -r requirements.txt

NVENC not available error

• Automatic fallback to libx264 (CPU)
• Make sure you have updated NVIDIA drivers

Audio not detected

• Verify the video has an audio track
• Try lowering the MIN_FREQ parameter

☕ Support the Project

If you find this tool useful for your content creation, please consider supporting its development!

🌟 Credits

If you use ASMR Pro Cutter for your videos, we'd love to see them! Feel free to mention "Made with ASMR Pro Cutter" in your video description to help others discover this tool.

📄 License

MIT License - See LICENSE for details

🤝 Contributing

Pull requests are welcome! For major changes, please open an issue first to discuss what you would like to change. See CONTRIBUTING.md for detailed guidelines.

👨‍💻 Author

Created with ❤️ by lordpba for the ASMR community

---

⭐ If you like this project, leave a star on GitHub!