Interview

About MXM

Max Møller makes the MXM collection alone, in Denmark. Max has made synth music for more than 40 years and has written software since the mid-1990s. This interview is from November 2026, edited for length.

Who is MXM?

Me, Max Møller. MXM is a one-person project, and a passion project. When I started making music I called myself Max M, and when I needed a website I wanted the shortest address I could get. MXM was short, and it reads the same backwards, which is the kind of joke programmers like.

Why build your own plugins?

Because I wanted to move to Linux, and for years I couldn’t. The music software wasn’t there. I would have had to give up a long list of the instruments and effects I use all the time.

So I started building what I was missing, with the help of AI. That makes the plugins opinionated. They do what I think matters when you make music, and they leave out what I don’t need.

The beta runs on Windows, macOS and Linux. The collection itself is now developed on Linux.

Where did it start for you?

With a Yamaha CS-40M, in 1983. It was duophonic, a smaller relative of the CS-80, and I hated it. I wanted to make synth pop like Depeche Mode, Yazoo and Kraftwerk, and it couldn’t make the sharp electronic sounds I was after.

I think my next synth was a Roland SH-101, and I loved it. It is the synth MXM Mono 01 is modelled on. Over the years I’ve had a lot of synths, a modular system among them.

How do you make a plugin that follows an old synth?

With research. I read the service manuals and the circuit schematics, and follow the signal path, the way the controls work on each other and the odd things the circuit does. I repaired TVs in my teens and I’ve been a software developer since about 1995, so I’m at home in both a schematic and the code.

The panel is the one thing I don’t copy. Each product page names the machine a plugin started from and what the two share.

What should the plugins sound like?

I like music that sounds as if it has a history and a place. The SH-101 has a clean, electronic sound, and I love that. But I also want to be able to make it sound as if it was sampled off an old record, or recorded somewhere in particular. So I have a soft spot for the old, grungy effects. I want to be able to make a sound dirty.

That’s why the collection has MXM FX Oldify for tape, vinyl and shellac, a bucket-brigade delay whose echoes wear as they repeat, a spring reverb, and a hundred rooms in MXM FX Convolution.

How do you know when it sounds right?

It’s a balance. An effect should be able to make a subtle change, and it should also be able to go all the way, into dirty noise or very electronic bleeps. So I check two things. The default settings have to sound good, and the range has to reach the extreme.

One example: everything with a delay or a reverb in it can go into feedback. But how far it can run away is limited, so it won’t blow your speakers.

Why don’t the plugins look like the hardware?

I can’t stand skeuomorphic design, where a plugin is drawn to look like a hardware panel. You leave so many possibilities on the table. A screen isn’t a panel. The controls can be grouped by what they do, the signal can be drawn so you see what is happening to it, and the layout can change with the window.

Every MXM editor is made of cards. Make the window smaller and the cards move onto pages, in an order you can predict, instead of shrinking. That matters on Linux with a tiling window manager: a plugin that opens in a small tile is as easy to edit as one in a big window.

Why learn yet another set of plugins?

Because you learn it once. The plugins share one way of working, so when you know one of them, you know the rest.

They can also be edited entirely from the computer keyboard. I’ve made a key language that is the same in every plugin. The arrows move between controls, three keys change a value in coarse, fine and finer steps, and F1 shows all the keys. You can make a sound without touching the mouse, and the DAW I’m building after 1.0 works the same way. Using the MXM editors has the details.

Why open source?

For longevity. The plugins are free software under the GPL, version 3 or later, so they won’t be abandoned as long as somebody uses them. Everything is documented well enough that a developer with an AI agent can pick it up and carry on. And for as long as I keep making music, I’ll maintain these plugins and make new ones.

The foundation they share, mxm-kit, is published under the MIT licence. It holds the editor, the presets, the key language and the shared signal processing. Anyone who builds a plugin for the collection starts with most of the decisions already made.

Anyone can build plugins with AI now. Why these?

That is what worries me most: that nobody will care when it’s done. Just as I found out I could build my dream plugins with AI, so did everybody else, and we’ll be flooded with plugins and DAWs.

What I bring is my taste, more than 40 years of making synth music, and one idea of how the plugins should work together. And even with AI, a plugin is months and months of work. It’s easier to use ones that are already made.

Still, I think the flood is a good thing. In the future, music software will be cheap, open source and on Linux. That matters more than whether people use my software in particular.

What would success look like?

If someone makes a good track with these plugins, I’d love to hear it. Send it to me, or post it on our Discord server.

I hope MXM can become my full-time work, perhaps with a few people to help it grow, and with an open-source community around it. The Discord server is where that starts.