Screen Recorder Pro

Windows screen recorder and screenshot tool built with Python, Tkinter and FFmpeg.

Screen Recorder Pro is a modular Windows desktop application for screen recording, screenshots, microphone/system-audio capture, webcam preview and detailed recording diagnostics. The current source build is 2026-08-12-screenshot-toolbar-persistence-v16.

The project prioritizes recording stability, useful diagnostics and safe cleanup of background processes. Recent work extends the screenshot workflow without rewriting the validated ddagrab/NVENC/CFR recording pipeline.

Highlights

• screen capture through FFmpeg Desktop Duplication (ddagrab) with GDI fallback;
• NVIDIA NVENC and CPU encoding options;
• microphone and Windows system-audio capture;
• CoreAudio loopback fallback when FFmpeg WASAPI is unavailable;
• global hotkeys, including native Windows RegisterHotKey handling for Print Screen;
• frozen-desktop region screenshots so menus and transient UI remain visible while selecting an area;
• screenshot toolbar with Region, Draw, Arrow, Undo and Clear tools;
• separate persistent colors and sizes for drawing and arrow tools;
• persistent screenshot-toolbar position with automatic clamping when monitor geometry changes;
• pause/resume and multi-segment finalization;
• webcam preview and annotation overlay;
• tray operation and Windows autostart support;
• managed FFmpeg/subprocess lifecycle and shutdown cleanup;
• structured diagnostic logs designed for root-cause analysis;
• automated structural/regression verification in verify_project.py.

Recording pipeline

The stable GPU path is:

Desktop Duplication / ddagrab
→ D3D11 frames in GPU memory
→ wall-clock timestamps
→ one CFR normalization step
→ NVIDIA NVENC
→ MP4 / MKV / AVI / MOV

For the project's validated 144 Hz → 72 FPS scenario, the application polls ddagrab at 144 FPS and produces a 72 FPS output while preserving wall-clock timing.

Project structure

main.py
Screen Recorder Pro.py
verify_project.py
Docs/
screen_recorder/
├── app.py
├── shared.py
├── components/
│   ├── annotation_overlay.py
│   ├── audio_loopback.py
│   └── webcam_preview.py
└── mixins/
    ├── audio_devices.py
    ├── capture_commands.py
    ├── finalize.py
    ├── instant_buffer.py
    ├── problem_logs.py
    ├── processes.py
    ├── recording_control.py
    ├── recording_session.py
    ├── screenshots_hotkeys.py
    ├── segment_audio.py
    ├── settings.py
    ├── smoothness_diagnostics.py
    ├── timing.py
    ├── tray_startup.py
    └── ui.py

ScreenRecorderProWin11 is assembled from mixins. The modules separate recording, audio, FFmpeg commands, lifecycle, UI, diagnostics, screenshots and finalization without changing the public launcher.

Detailed architecture and troubleshooting notes are available in Docs/README.md.

Requirements

• Windows 10/11
• Python
• FFmpeg and FFprobe available through PATH
• Python dependencies from requirements.txt
• NVIDIA GPU is recommended for the NVENC path; CPU encoders are also exposed by the application

Install Python dependencies:

python -m pip install -r requirements.txt

Run:

python "Screen Recorder Pro.py"

or:

python main.py

Verification

The repository includes a project verifier:

python verify_project.py
python -m compileall .

verify_project.py checks project structure, required methods, system-audio selection regressions, screenshot-region behavior and safe log-cleanup ownership rules.

The v16 source package was checked before publication in a non-Windows environment: verify_project.py completed successfully and compileall found no Python syntax errors.

Hardware-dependent behavior — real ddagrab, NVENC, Windows clipboard, native hotkeys and physical audio devices — still requires verification on Windows hardware.

Diagnostics and reliability

The application contains dedicated diagnostics for FFmpeg commands, recording events, timing, smoothness, screenshot selection, hotkeys, errors and source-version fingerprints. Runtime logs, settings, temporary recordings and generated media are excluded from Git.

Long-running and external-process code is designed around:

• managed child processes;
• explicit shutdown cleanup;
• timeouts and return-code checks;
• bounded diagnostic files;
• temporary-file cleanup;
• preserving useful recording output when recovery is safe.

AI-assisted development

AGENTS.md and the Docs/ folder document project invariants, architecture and validation rules used during AI-assisted development with Codex/ChatGPT. AI-generated changes are treated as proposals and are validated with repository checks and manual Windows testing where hardware behavior is involved.

License

No open-source license is currently granted. The repository is public for portfolio and source-review purposes.