What if every track in your Logic Pro session could talk back to you? What if you could tell your kick drum to “turn down 3 dB” or ask your bass to solo itself—all while your song keeps playing? I’m excited to introduce Chatty Channels, an open-source project that transforms Logic Pro into an AI-powered collaborative studio where natural language meets professional music production.

Beyond AI Music Generation: True Studio Collaboration

Unlike AI tools that generate music from scratch, Chatty Channels enhances your existing creative workflow. Picture this: you’re in the middle of a mix, and instead of reaching for your mouse, you simply say “add more reverb to the lead vocal” or “bass, can you solo yourself?” The system responds instantly, making precise adjustments while preserving your artistic vision.

This isn’t just automation—it’s collaboration. By placing lightweight AI-powered plugins on each channel, Chatty Channels creates virtual band members that understand their instruments and respond to nuanced direction. The master bus hosts an AI engineer with years of mixing expertise, while a beautiful Control Room app serves as your command center, complete with vintage TEAC-style VU meters and intuitive chat interface.

The Technology That Makes Magic Possible

Behind the scenes, Chatty Channels showcases some impressive technical achievements. The system uses ultra-low latency OSC (Open Sound Control) communication, achieving round-trip times under 200ms—fast enough to feel instant during creative flow. A sophisticated PID control loop ensures your adjustments are precise, typically hitting target values within 2-3 iterations.

The Control Room supports multiple AI providers seamlessly—OpenAI, Google’s Gemini, Anthropic’s Claude, and xAI’s Grok—so you can choose the AI that best matches your working style. The system even includes automatic calibration that identifies which plugin lives on which track, eliminating setup headaches.

Built by Musicians, for Musicians

Drawing from my two decades in open-source development and 10 years as a sound engineer, I designed Chatty Channels to enhance rather than replace human creativity. The vision is simple: create a virtual collaborative studio where technology amplifies your musical ideas.

The architecture reflects real studio hierarchy. AI Musicians on each channel understand their instruments intimately. The AI Engineer on the master bus provides mixing guidance and technical expertise. And you, the Producer in the Control Room, orchestrate everything through natural conversation—just like directing a real session.

Open Source: Built in the Open, for the Community

Following the success of my previous open-source project GPSTracker (over 2.2 million downloads since 2007), Chatty Channels is being developed completely in the open. The codebase features production-grade quality with comprehensive documentation, automated CI testing, and a modular SwiftUI and JUCE architecture that makes it easy for contributors to add new AI personas, effects, or DAW integrations.

This isn’t just about sharing code—it’s about building a community. Whether you’re a developer interested in AI integration, a producer wanting to shape the future of music technology, or simply curious about what’s possible, there’s a place for you in this project.

Current Progress and What’s Next

The development has been moving fast. Version 0.5 proved the core concept with successful closed-loop control of track parameters. Version 0.6 added those gorgeous vintage VU meters and multi-LLM provider support. I’m currently deep in version 0.7, focusing on real-time data visualization and bulletproof OSC message handling.

The roadmap ahead includes lazy FFT calculations for efficient frequency analysis, strict JSON schema validation for AI responses, and ultimately a public alpha release that will put this technology in the hands of music producers worldwide.

Join the Future of Music Production

Whether you want to contribute code, test features, suggest improvements, or simply watch this technology evolve, I’d love to have you aboard. This is more than just another music tool—it’s a glimpse into a future where AI and human creativity work together to push the boundaries of what’s possible in music production.

The studio of tomorrow is being built today. Come help create it.

It’s been almost 3 years since I’ve posted anything on this website. Gpstracker is still going strong. It just keeps on running without my intervention. And people are still downloading it, which is a good thing.

I retired from writing software 5 years ago and since then I’ve been practicing my guitar and learning music theory and I’ve written my first song! I sang the vocals and played guitar. I used logic pro for the drums and ezbass for the bass. The name of the song is “love, love, love”. Check it out and let me know what you think!