VideoRefurbish (PyQt5) — 3GP/Old-Phone Video Cleanup GUI

This is a beginner-friendly PyQt5 GUI that uses FFmpeg to:

• convert old .3gp clips to .mp4
• reduce blockiness / noise (common on early phone cameras)
• upscale with Lanczos (optional)
• clean up audio (denoise + basic EQ + normalize)

It is NOT “magic AI upscaling” (though you can get surprisingly decent results).

What you need

• Windows 10/11 (works on Linux too, but the “portable FFmpeg download” button is Windows-focused)
• Python 3.10+ recommended
• FFmpeg (either installed in PATH or downloaded via the GUI button)

Quick start (Windows)

1. Unzip this folder somewhere (e.g. D:\VideoRefurbish)
2. Run: scripts\install_windows.bat
3. Run: scripts\run_windows.bat

Linux quick start

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python src/refurbish_gui.py

Install ffmpeg via your package manager (e.g. sudo apt install ffmpeg).

Notes

• The GUI shows the exact FFmpeg command it will run (copyable).
• Progress is based on ffmpeg -progress pipe:1 + ffprobe duration.
• Output is encoded as H.264/AAC (broad compatibility).

Typical tips for very old 3GP material

• Start with Preset: Balanced.
• If it’s extremely blocky: enable Deblock + Denoise (strong).
• If it becomes too “waxy”: reduce denoise or disable sharpen.
• For audio that’s “telephone-like”: enable Voice Cleanup and keep the lowpass around 7–9 kHz.

Have fun restoring your time-capsule footage 🙂