Video Uniqueizer

Fast, visually lossless tool for automating video uniqueization for Instagram Reels. Bypasses duplicate-detection algorithms by altering video and audio fingerprints while keeping pristine picture quality intact.

Key Features

• Micro-Zoom (100–104%): Shifts the pixel grid without blurring or cropping out important content.
• Micro-Speed (98–102%): Alters video duration and audio waveform while preserving original audio pitch.
• Micro-Color Grading: Subtle brightness, contrast, and saturation shifts (±1.5%) that keep skin tones natural.
• Metadata Stripping: Wipes all camera EXIF data, creation timestamps, and encoding tags (-map_metadata -1).
• Blurred Background: Automatically pads horizontal videos into 9:16 vertical format (1080x1920).
• GPU Accelerated: Uses hardware NVIDIA NVENC (h264_nvenc) — renders a 30s 60fps video in ~3-4 seconds. Falls back to CPU (libx264) if no NVIDIA GPU is detected.

Modes

1. Multi-Account Batch Mode (scripts/run_batch_x12.bat)

Generates 72 distinct unique videos and organizes them directly into separate folders for 12 Instagram Reels accounts (6 unique videos per account):

output/
  ├── account_01/  ->  6 unique videos
  ├── account_02/  ->  6 unique videos
  ...
  └── account_12/  ->  6 unique videos

Every single file across all 12 accounts receives its own random seed and parameters, ensuring zero duplicate collision between accounts.

2. GUI Mode (scripts/run_gui.bat)

A clean PyQt5 interface with tabs for:

• 🎬 Уникализация видео: custom one-off video processing with manual controls over filters, zoom ranges, speed ranges, and copy counts.
• 🎵 Уникализация аудиодорожки: extracts voiceover/audio, modifies acoustic fingerprint (pitch ±1.8%, tempo ±0.7%, formant preserved, dynamic EQ), and exports MP3 192k + iOS Camera Roll MP4 containers.

3. Standalone Audio Mode (scripts/run_audio_unique.bat)

Fast batch uniqueization of voice/audio tracks directly from the command line or via interactive drag-and-drop:

• Input: Any video or audio file
• Output: N unique MP3 (192 kbps) + N iOS Camera Roll containers (.mp4) with zero metadata.

Quick Start

1. Drop your source video(s) into input/.
2. Run:
   • scripts/run_audio_unique.bat (to uniqueize voiceover/sound tracks)
   • scripts/run_batch_x12.bat (for multi-account batch rendering)
   • scripts/run_gui.bat (for full GUI control)
3. Grab the generated files from output/.

Requirements

• Windows 10/11
• Python 3.10+ (pip install -r requirements.txt)
• FFmpeg installed and available in system PATH
• (Optional) NVIDIA GPU for fast NVENC hardware encoding