Within Open Culture Tech we're building a free toolkit that artists can use without paying a cent, and that stays the foundation of the project. But building a tool costs money, and maintaining it costs at least as much, so at some point you arrive at the question you'd rather avoid: who actually pays for this.

 

Over the past year and a half we looked into three routes. The first was about licensing, and it gave us the clearest answer. The toolkit stays free for everyone in the music sector, but a company that wants to use the tool commercially outside that sector pays for it through a separate license. The artist never notices any of this. Our point isn't that artists should never pay for anything, it's that the bill for commercial development shouldn't land on them.

 

While working this model out, we also looked at how other open source projects in the music world have solved this, because we're certainly not the first to wrestle with it. Ardour, a free and open source DAW, does something similar. The source code is and stays completely free, anyone can download it and build it themselves. But anyone who doesn't feel like building it and just wants a working version pays for that, either through a subscription or a one-time payment. That lines up almost exactly with what we're trying to do with our license route: the core stays free, convenience costs money. The difference is that Ardour sells that convenience to individual users, while we aim it at companies that want to use the tool commercially.

 

Sponsoring and paid workshops are still much less settled. We've had conversations with companies and funds that showed interest, without anything being signed. With the workshops we kept running into the same tension: how do you charge money for something you're fighting just as hard to keep free for the artists who need it.

 

Here too it helped to look at how others handle this. Blender, the open source 3D software, funds most of its core development through a Development Fund that companies join as sponsors, with contributions running into the tens of thousands of euros per year. What's good about it is that a company that pays in gets no say over what gets built. Sponsorship buys continuity, not influence. That fits well with our own values around transparency and autonomy, and it's exactly the kind of condition we should lock in for our own sponsorship before we really get started with it. Blender combines this with paid trainer certification and a separate subscription for extra training material, while the software itself stays free.

 

For the workshops we looked closer to home, at Waag Futurelab in Amsterdam, an organization that, like us, works from public values. Waag keeps its own workshops and events largely free or "pay as you can," and generates income through paid research commissions for municipalities and other public institutions. The question that raises: instead of selling workshops directly to record labels, could we look at paid commissions from institutions that need research or expertise in this area?

 

There's one more route we'd overlooked until now, and it actually fits our own tech partners best. The NLnet Foundation manages the NGI Zero programme on behalf of the European Commission, specifically meant to fund open source software, open hardware, and open standards, with grants between 5,000 and 50,000 euros per project. This isn't sponsorship from a company or a commercial license, but public money specifically earmarked for exactly the kind of work our tech partners already do. Instead of a tech partner having to commercialize their work to get anything out of it, they could turn directly to funds like this, as long as the result stays open source.

 

What struck us when we set these examples next to our own routes: none of them simply transfers over. Ardour sells convenience to individuals, which is exactly what we don't want to do to artists. Blender's sponsorship works mainly because its community is so large that companies are literally dependent on the platform, and we don't have that scale. Waag's commission model works well for research, but less so for something tangible like a tool or a show. NLnet is perhaps the most directly applicable example, because it lines up precisely with what our tech partners already make. What we take from all four is the underlying principle: as soon as money comes in through a party that benefits itself, such as a company, a government, a fund, or a paying user, rather than through the artist, accessibility stays intact.

 

This question isn't just ours. Artists, the tech partners we build with, and the venues where it all eventually gets played, run into the same problem. What follows are ideas we haven't tested yet, but that we think are worth sharing.

 

For artists we're thinking of something like what Blender does with its training subscription. The basic workshop should stay free for everyone, but anyone who wants to go deeper after that workshop, with someone who already knows the tool well, could pay for that time. Another thought, closer to Blender's certification model, is that artists who've mastered the tool could become a kind of recognized guide for other artists, for a small fee.

 

For the tech partners we work with, the license route makes the most sense, exactly as with Ardour. It would also be worth exploring whether funds like NLnet could be a structural addition to what we're doing now, precisely because they're specifically meant for this kind of work.

 

For venues we're still searching the most, and Waag's model seems like the most usable starting point here. A venue doesn't build anything and doesn't use the tool itself, but it offers the space where it all happens. Instead of passing on costs to the artist or the audience, a venue could look at paid partnerships with funds or municipalities interested in this kind of programming.

 

None of this is settled yet. These are ideas from a year and a half in which we were mostly searching for what does and doesn't work, sharpened by looking at how others in comparable open source projects have tackled the same problem.