Video flip fixer

---

Application built with Python 3 and PyQt6. This tool performs lossless (no re-encoding) resetting of video orientation metadata matrices (such as the Apple iPhone -display_rotation flag) in just a fraction of a second.

It uses an advanced Web Deployment / Bundled Hybrid Architecture powered by static-ffmpeg to achieve complete autonomy and offline capability without requiring any pre-installed global system dependencies.

Key Features

• ** 100% Lossless & Instant:** Resets hidden rotation metadata flags (0°) while perfectly copying original video and audio streams (-c copy). Processing a 500MB file takes less than 0.5 seconds.
• ** Zero-Dependency Executable:** Bundles full-production ffmpeg and ffprobe binaries internally within a single, portable Windows .exe application via cloud automation.
• ** Robust Offline Capability:** Once compiled or deployed, the application operates in a completely isolated air-gapped environment. It never makes stealthy update checks or network calls during runtime.
• ** Professional Camera Support:** Enhanced block-by-block streaming metadata parser securely resolves complex or non-standard index definitions (moov / tkhd atoms) found at the very end of heavy cinema files (Sony, RED, ARRI, iPhone 15 Pro+ ProRES).
• ** Automated CI/CD Releases:** Fully integrated with GitHub Actions. Pushing a version tag automatically compiles, signs, and attaches the release asset to your repository.

---

Core Component Architectures

1. Graphical Interface (main.py)

Implements an asynchronous PyQt6 thread workflow (QThread). Heavy operations (file IO execution, metadata extraction, structural cloning) are completely decoupled from the main thread, keeping the GUI smoothly responsive and preventing accidental OS freezes during batch operations.

2. Video Processing Engine (core.py)

Bypasses restrictive wrapper abstractions by executing clean, low-level subprocess operations with CREATE_NO_WINDOW flags to suppress console popups. It dynamically maps execution directories using sys._MEIPASS when frozen inside a virtual boot sandbox environment.

3. Cloud Packaging Specification (build_release.py)

Automates the local discovery of absolute module environment paths on virtual hosts. It structures the output parameters of PyInstaller using the --add-data protocol, mapping compiled binary payloads directly into an internal standalone application subfolder named bin/.

---

Local Development & Setup

To run the project locally from source using Visual Studio Code, follow these environment installation steps:

1. Clone the repository:

   git clone https://github.com
   cd video-flip-fixer

2. Install Python environment dependencies:

   pip install PyQt6 requests static-ffmpeg pyinstaller

3. Verify engine integrity: Run the main entry script directly from your terminal:

   python main.py

   Note: On your first local launch, static-ffmpeg will securely stream verified, official production binaries into your Python site-packages environment layout.

---

CI/CD Deployment via GitHub Actions

The compilation process is fully handled in the cloud using Microsoft Windows Runners. You never need to manually pack your files or risk local character mapping (CP1252 vs UTF-8) compilation crashes.

To release a new version with an automated version-tagged executable (e.g., video-flip-fixer-v1.0.0.exe):

1. Commit and push your clean workspace changes:

   git add .
   git commit -m "Release stable production core build"
   git push origin main

2. Create and publish a version tag:

   git tag -a v1.0.0 -m "Version 1.0.0 Release"
   git push origin v1.0.0

3. Monitor Cloud Pipeline: Navigate to the Actions tab on your GitHub repository page. The automated runner will spin up a fresh Windows environment, resolve execution configurations, compile the single binary payload via PyInstaller, and attach the portable application to a freshly minted GitHub Release page automatically.

---

Technical Specifications

• UI Framework: PyQt6 (Qt 6.x)
• Binary Stream Core: FFmpeg / FFprobe (Full Essentials Build Layer)
• Compilation Engine: PyInstaller 6.x+ (Configured with custom system stream interceptors to block windowed environment crashes)
• Target OS Deployment: Windows 10 / 11 (Architecture target x64)
• Metadata Target: tkhd atom display transformation matrix coordinates (Resetting translation maps to uniform identity grids)

License

This repository is distributed under the terms of the MIT License. You are free to copy, modify, distribute, or use this extension setup commercially, provided the original author credits remain intact. See the LICENSE file for details.

Support the Project

If this plugin saves you time and helps you keep track of your calculations, consider supporting its further development! Any contribution is highly appreciated.

💵 USDT (TRC-20): TS1x5AR87CipqAqF3bb7sqePBKqRLFEc4Q

🪙 Bitcoin (BTC): bc1qpamcvn5le907jkt5zhhd4c2r783489nts72m8m

💎 Ethereum (ETH): 0x354dc4b9cEC4fC05021Cd26734B8cc839390e5af