The latest articles on DEV Community by David Chen (@lucemia). https://dev.to/lucemia https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3219084%2Ff8c3bd68-8a18-4f2f-89f7-d8988ee9624b.jpeg https://dev.to/lucemia en David Chen Wed, 28 May 2025 17:54:00 +0000 https://dev.to/lucemia/typed-ffmpeg-30-strongly-typed-ffmpeg-interface-for-python-with-visual-playground-1ep9 https://dev.to/lucemia/typed-ffmpeg-30-strongly-typed-ffmpeg-interface-for-python-with-visual-playground-1ep9 <p>🎬 typed-ffmpeg 3.0 – A Strongly-Typed Python Interface to FFmpeg + Interactive Playground</p> <p>If you’ve ever written long, fragile FFmpeg command strings and wished for a better developer experience, typed-ffmpeg might be what you’re looking for.</p> <p>It brings autocomplete, type safety, validation, and now even a visual editor to FFmpeg workflows—all from Python.</p> <p>GitHub: <a href="https://github.com/livingbio/typed-ffmpeg" rel="noopener noreferrer">https://github.com/livingbio/typed-ffmpeg</a></p> <p>⸻</p> <p>🚀 What typed-ffmpeg Does</p> <p>typed-ffmpeg lets you build FFmpeg filter graphs in Python using a typed, object-oriented API. It solves common issues with tools like ffmpeg-python, such as: Missing type hints and autocomplete</p> <p>It also introduces:</p> <ul> <li>JSON serialization of graphs</li> <li>FFmpeg command validation and auto-correction</li> <li>Reverse CLI parsing (in v3.0)</li> </ul> <p>⸻</p> <p>👥 Who It’s For</p> <p>If you work with video/audio processing and want to:</p> <ul> <li>Automate FFmpeg workflows in production</li> <li>Build content/media pipelines for research or education</li> <li>Debug and share filter graphs more easily</li> </ul> <p>⸻</p> <p>🔄 What’s New in v3.0</p> <p>Here are the highlights from the latest release:</p> <p>✅ Source Filter Support</p> <p>Use FFmpeg source filters like color, anullsrc, testsrc, etc., with full autocomplete and type safety.</p> <p>Docs: <a href="https://livingbio.github.io/typed-ffmpeg/usage/complex-filtering/#applying-source-filters" rel="noopener noreferrer">https://livingbio.github.io/typed-ffmpeg/usage/complex-filtering/#applying-source-filters</a></p> <p>✅ Input Stream Selection</p> <p>Now supports stream specifiers (e.g., [0:v], [1:a]) across multiple inputs.</p> <p>Docs: <a href="https://livingbio.github.io/typed-ffmpeg/usage/basic-api-usage/#stream-selection" rel="noopener noreferrer">https://livingbio.github.io/typed-ffmpeg/usage/basic-api-usage/#stream-selection</a></p> <p>🧪 Interactive Playground</p> <p>A new visual tool to build, edit, and test FFmpeg filter graphs:</p> <p>🔗 <a href="https://livingbio.github.io/typed-ffmpeg-playground/" rel="noopener noreferrer">https://livingbio.github.io/typed-ffmpeg-playground/</a></p> <p>Features:</p> <ul> <li>Drag-and-drop filter graph builder</li> <li>Generates both FFmpeg CLI and typed-ffmpeg Python code</li> <li>Paste any FFmpeg CLI to reverse-parse it into a visual graph</li> </ul> <p>🧱 Internal APIs for Tooling</p> <p>v3.0 also introduces:</p> <ul> <li>parse cli to graph: – Convert FFmpeg CLI to internal graph objects</li> <li>generate python code: – Emit typed-ffmpeg code from a graph</li> </ul> <p>These make it easier to build advanced tooling or plugins around FFmpeg logic.</p> <p>⸻</p> <p>📦 Installation</p> <p>pip install -U typed-ffmpeg</p> python ffmpeg typehint