阅读 中文版.
Check out the video demonstration
This is a Python-based Rotaeno screen recording stabilization script. Its principle is the same as the script provided by Rotaeno for Adobe After Effects, which rotates frames based on the colors of the four corners of the screen recording in live broadcast mode. The goal is to stabilize the video image.
- Added background circle, optimized visual experience
- Fixed some bugs when the script runs on Windows
- Fixed the issue where the background circle could not be properly added to screen recordings with a small width-height ratio.
- Added support for mov, avi, mkv, wmv, flv formats
- Added multi-process optimization.
- Added square rendering feature (thanks to the PR by @Ki-wimon). The script now defaults to square rendering to minimize cropping of the frame.
- Added a
verbose=Falseparameter to theconvert_vfr_to_cfrandadd_audio_to_videofunctions to reduce redundant command line output. - Deleted intermediate files, keeping only the final output.
- Added the V2 correction method. The script now defaults to stabilizing videos using V2. If V1 correction is needed,
please add "v1" at the beginning of the video filename. The script will automatically switch to V1 correction mode for
stabilization, e.g.,
v1-sample.mp4.
- No need to install Adobe After Effects; rendering can be done with just one command.
- Supports batch processing of videos.
- Download the project code:
git clone https://github.com/Lawrenceeeeeeee/python_rotaeno_stabilizer.gitAlternatively, you can directly click "Download ZIP" on this repository page, then unzip the downloaded file.
- Install the dependencies:
# Navigate to the directory containing the script
cd python_rotaeno_stabilizer
pip install -r requirements.txt- Install FFmpeg
Please download the appropriate installation package from the FFmpeg official website.
-
Attention! Before recording, ensure the "Streaming Mode" is activated in Rotaeno settings. Once enabled, the four corners of the screen will display color blocks, which indicate the device's rotation angle.
-
Place the video to be processed in the videos directory (currently supports mp4, mov, avi, mkv, wmv, flv).
-
Start the project:
python main.py- Once processing is complete, find the rendered videos in the
outputfolder.
Result as follows:

