I just wanted to show you guys a little bit of behind the scenes. I’m diligently working. The chromesthesia will read your mix and change colors to show you where the frequencies lie. It’s pretty neat. I just wanted to share it with the group and let y’all know the beta will be coming.

Man, I did NOT expect this — thank you. Reading every single comment. 🙏🔥 A lot of y'all are asking where to sign up for the beta — it's coming, and I'm pulling the first testers from the free community (link in the description). Early access + the behind-the-scenes builds drop there first. One question — and your answer literally decides what I point MIXSTEIN at next: 👉 What part of mixing kills your momentum the most — GAIN STAGING, DE-ESSING, or BALANCING FADERS? Drop it below.

I'm not a coder. But over the last few months I've quietly built an AI system that runs most of my business — and every time someone sends me a "this is the new way to work with AI" video, I watch it and realize I'm already past it. Here's the part nobody says out loud: the people actually winning with AI aren't the ones who hand it the wheel. They're the ones who keep their taste in the driver's seat. AI is the workforce. Your ear is still the boss. That's the whole philosophy — and it's the same reason a great mix was never about the plugins. A few principles I run on: 1. Taste stays in charge. AI does the work; I make the calls. The second you outsource the judgment, you're just average — faster. 2. Point it, don't flood it. Same as soloing to find one frequency: give it the one thing that matters, not everything at once. 3. When a move works, I save it. Repeatable every time. I never build process for the sake of process. 4. Nothing ships on the first pass. I make it check its own work against a hard standard until it's actually right — like A/B-ing to a reference until it translates. 5. The frameworks I'll teach all day. My actual taste stays mine. I've been doing all this in the dark. Starting to think that's the mistake. What's the one task in your workflow you'd hand to AI tomorrow — if you trusted it not to ruin your sound?

If you came in off the video — no, that wasn't clickbait. MIXSTEIN is real, it runs in my actual sessions, and you just walked into the room where it gets built in public. Quick on me: 20+ years mixing and mastering, real outboard — Neve 5059, SSL Fusion, the WesAudio chain — and I'm building MIXSTEIN because every "AI mixing" tool I've tried treats your song like a preset. MIXSTEIN isn't that. It's Mixing Intelligence, not AI — an instrument that learns you, not the internet's average. What this room is: producers and engineers sharpening craft out loud. Real gear, real sessions, real decisions — and the why behind them. I post the build as it happens (the wins and the stuff that breaks), break down real mix moves, and you get first crack at the MIXSTEIN beta before anyone outside these walls. What it's not: a place to lurk. The people who get the most here are the ones who post. So let's start the first real thread 👇 What kills your mix momentum? The exact moment a session stalls — vocals won't sit, low end turns to mud, last night's mix sounds flat this morning. Be specific. That stall point is exactly what I'm building MIXSTEIN to kill — and your answer shapes what I build next. Welcome in. 🎛️