Every piece begins long before any code is written. An idea for a new algorithm can live in my head for months: I carry it around, construct it mentally, trace its lines in my mind. Only when it holds up there do I write a small Python prototype to see whether the method works at all.

Then the slow part starts. I refine the code and add parameters to control the behavior in finer detail, usually somewhere between ten and thirty, over several weeks. My own tooling turns the result directly into HP-GL, G-code or SVG, including path optimizations for the specific machine. After that, the experiments move to the plotter: speeds, pens, pen pressure. I keep records of every run, and the practical knowledge about each machine keeps growing. Some algorithms need hours, sometimes a full day, to compute a single image. I let them run: whether something interesting emerges often only shows over that kind of time. Optimizing the code comes later, once I am sure the path is worth following.

Watching a plotter draw is my version of staring into a campfire: the image builds up line by line, and I cannot look away

Most of my machines are A3 flatbeds; for those I prefer bristol board. The large-format plotters sometimes run on original paper rolls from the eighties. I draw mostly in black, with technical ink pens and vintage plotter pens, some of which I have modified. The vintage felt tips in particular are a challenge after forty years, but when they work, they put something on the paper that no modern pen does. I like it when the pen leaves its character in the image.

Some motifs exist as small series of five to twenty sheets, signed and numbered, no two of them identical, because every pen draws a little differently. Other pieces are plotted exactly once. There is nothing digital for sale: the work is ink on paper.

I work in Hamburg and am open to commissions, collaborations and exhibition inquiries. If you have something in mind, get in touch.