The story
Now in your browser. And in your ears at 3am.
Carmina.audio is a solo project. One engineer, no team, no investors, no roadmap driven by a board.
I've spent 20 years in live and studio audio — mixing concerts, recording albums, designing IEM systems, fighting low-end masking on stage monitors at 1am with a show in 8 hours. I know what a mix sounds like when it's wrong, and I know exactly why it's wrong.
I built Mixtus because I got tired of seeing producers upload their mixes to forums and get back "sounds good!" or "maybe boost the highs a bit." That's not feedback. Feedback is: "your kick and bass are fighting at 80Hz, your integrated LUFS is 3dB too hot for Spotify, and your vocals are collapsing to mono in the 2–4kHz range." That's what an experienced engineer hears in the first 90 seconds.
I built Otium because I have tinnitus. I know what it's like to lie awake at 3am with a 6kHz tone that won't stop, trying every app on the App Store and finding looped rain sounds that your brain learns to ignore within a week. I built what I actually needed.
How we build
Carmina.audio runs on a VPS in France. Not AWS. Not Google Cloud. This means your audio files are processed on infrastructure I control, deleted within the hour, and never fed into any third-party machine learning pipeline. The Mixtus analysis API is open source — you can read every line of code that touches your mix.
There's no seed round, no Series A, no growth targets to hit. I'm building tools I'd use myself, at a pace that doesn't compromise quality. If a feature isn't right, it doesn't ship. This is sustainable by design — not by accident.
Otium collects nothing. Mixtus processes your file and deletes it. There are no trackers, no analytics scripts, no cookie banners because there's nothing to consent to. This site loads faster and respects you more because of it.
The products launch in English — the market is global, audio is a universal language. But the company is French, the infrastructure is French, and the privacy principles follow French and European standards. SIRET registered, SAS structure, accountable.
The suite
AI mix feedback. Upload your track, get specific engineer-grade observations in under 60 seconds.
Sound therapy for tinnitus, misophonia and hyperacusis. Infinite non-repeating streams.
IEM coach for live musicians. Guided setup for Behringer P16, Roland M-48, Aviom and more.
Hearing health and noise dose meter for gigging musicians and live engineers.
Questions about the tools, feedback on the analysis, a bug to report, or just want to talk audio? I read every email.
hello@carmina.audio →