Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/kerasty2024/smart_encoder
Smart Encoder is a powerful tool for video encoding that optimizes video quality while minimizing file size. It automatically selects the best encoding parameters based on video content and desired output characteristics.
https://github.com/kerasty2024/smart_encoder
ab-av1 encode ffmpeg python
Last synced: 3 months ago
JSON representation
Smart Encoder is a powerful tool for video encoding that optimizes video quality while minimizing file size. It automatically selects the best encoding parameters based on video content and desired output characteristics.
- Host: GitHub
- URL: https://github.com/kerasty2024/smart_encoder
- Owner: kerasty2024
- Created: 2024-08-06T09:37:18.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-10-17T01:12:56.000Z (4 months ago)
- Last Synced: 2024-10-19T02:35:46.847Z (4 months ago)
- Topics: ab-av1, encode, ffmpeg, python
- Language: Python
- Homepage:
- Size: 390 KB
- Stars: 11
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Smart Encoder
Smart Encoder is a powerful tool for video encoding that optimizes video quality while minimizing file size. It
automatically selects the best encoding parameters based on video content and desired output characteristics.## Features
- **Automatic Codec Selection**: Automatically selects the best video codec for encoding (default: libsvtav1).
- **CRF Optimization**: Finds the optimal Constant Rate Factor (CRF) for the best quality-to-size ratio by ab-av1. If
suitable crf cannot be defined, use MANUAL_CRF(default: 23, usually recognized as visually lossless). If encoded file
size > original file size, increment CRF by MANUAL_CRF_INCREMENT_PERCENT (default: 15%) and encode again, until
encoded file size is smaller than original one.
- **Multi-Process Encoding**: Supports encoding multiple files in parallel to speed up the process. # of processors can
be configured by command-line.
- **Comprehensive Logging**: Logs success and error details for easy troubleshooting and monitoring.
- **Flexible Configuration**: Customizable settings for advanced users through configuration files.## Requirements
- Python 3.7+
- [FFmpeg](https://ffmpeg.org/download.html) installed and added to PATH
- [ab-av1](https://github.com/alexheretic/ab-av1/releases/) added to PATH for searching best codec and CRF.## Installation
1. Clone the repository:
```bash
git clone https://github.com/kerasty2024/smart_encoder.git
cd smart_encoder2. Create a virtual environment and activate it:
```bash
python -m venv venv
venv/bin/activate # On Windows use `venv\Scripts\activate`3. Install the required packages:
```bash
pip install -r requirements.txt4. Ensure FFmpeg & ab-av1 and so on (refer to Requirements) is installed and accessible from your command line.
## Usage
1. Prepare your media files: Place the media files you want to encode in a source directory.
2. Run the encoder:```bash
python -O main.py --processes 3 --move-raw-file # you can also modify # of processes, depending on how powerful your environmet is.- Use --manual-mode if you wish to manually set encoding parameters. manual crf can be set in settings/video.py
```bash
python -O main.py --processes 1 --move-raw-file --manual-mode # you can also modify # of processes, depending on how powerful your environmet is.
- If you want more message shown in console, remove -O option (change to debug mode).
- other options can be found in main.py3. Check logs:
- Success logs are stored in the output directory with details of encoded files.
- Error logs are available in the same directory for any issues encountered during encoding.## Configuration
.py files in settings directory allow you to adjust default settings for encoding parameters and language detection
preferences. Customize this file to suit your needs before running the encoder.## Contributing
Contributions are welcome! If you have ideas for improvements or new features, feel free to open an issue or submit a
pull request.## License
This project is licensed under the MIT License.