EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Jul 30·5 min read
Show HN: I think I made my own web version of minimalistic audacity
TL;DR
HERTZ is a web-based Digital Audio Workstation that enables real-time vocal processing directly in the browser, eliminating the need for server-side processing. It utilizes the Web Audio API and FFmpeg.wasm to provide a high-fidelity audio editing experience with minimal latency and privacy concerns.
✦ Why It Matters
Try HERTZ today for efficient, privacy-focused audio processing directly in your browser.
Key Takeaways
- Explore HERTZ for client-side audio editing without server dependencies.
- Utilize the specific vocal chain settings for optimal podcast clarity.
- Ensure your web server supports required security headers for HERTZ.
- Clone the repository and run HERTZ locally for immediate use.
How It Works
HERTZ processes audio using the Web Audio API, which allows for real-time manipulation of audio signals directly in the browser. It decodes various audio formats into a raw AudioBuffer and applies a vocal processing chain that includes equalization, compression, and normalization.
The use of FFmpeg.wasm enables encoding without server interaction, while SharedArrayBuffer facilitates efficient multi-threaded processing.