Harry van Haaren, open source musician

Harry van Haaren started releasing Linux audio apps in 2013, each one funded by a time-released donation model. His brand "OpenAV Productions" has added instruments and effects targeting electronic music production, together with a minimalist standalone workstation called Luppp, tailored for live looping. We dragged him away from wall-shaking beatbox rehearsals at the Linux Audio Conference in Karlsruhe, Germany, to investigate the past and future of his work producing popular music on Linux.

Sam Tuke: What audio apps are you working on at the moment?

Harry van Haaren: Luppp (a live looping program), ArtyFX (a suite of audio-effect plugins), and various other instrument plugins including Fabla (a drum sampler) and Sorcer (wavetable synthesizer). OpenAV is also working on an app for busking musicians and accompaniment (stay tuned!).

ST: What is "live looping"?

HvH: When you record a music sample and play it back on repeat – that's looping. It's become a popular way to make music in a variety of genres, electronic music most of all. You can use a mix of pre-recorded samples and instruments, including acoustic instruments and vocals, and use looping to create many layers of sound.

Live looping is when the process is performed in real time – the arrangement and playback is managed interactively as the music plays. This can be either for an audience at a concert, or for creation of a new recording.

ST: How does your workstation app, Luppp, improve on others available for Linux?

HvH: The apps I work on are designed to complement a live looping workflow: They focus on letting people create music on the fly and improvise. They're designed for live performance rather than studio use, but can be used for either.

In the Linux realm some looping apps, or "loopers" already exist, but Luppp has more organizational features: scenes and more high-level concepts to control the music session you're working on. These features don't exist in any other Free Software app. It's a question of workflow – how you work with the computer, how you organize sound samples, and how you create the loops. When you turn that knob, you should instantly get the change you want, and not have to click through menus and dialog windows.

ST: Your audio app interfaces all share a distinctive look – like curvy orange neon. Are you modeling them on proprietary apps?

HvH: Of course OpenAV takes inspiration from other software, both proprietary and open source. The user interfaces are designed particularly for live-performance workflow, and I use colors to highlight the important feedback. In general, I adhere to interaction design standards and use a consistent look and feel to ensure users are comfortable. Once a user interface is almost completed, it's tested and reviewed by beta-testers to ensure it's easy to understand.

ST: When did Luppp development start?

HvH: I've only been programming for 5 or 6 years with C++ and real-time audio. I began Luppp quite early – there are earlier versions written in Python and C++ that I never released. I learned all the difficult concepts of real-time audio programming like that, teaching myself. Although it continues to be an educational process, I'm confident the foundation of Luppp is now implemented in a robust and scalable way.

ST: So, like many hackers, you learned by doing, and Luppp was the app that you learned with?

HvH: Yes, very much so, and there were many occasions when I'd keep the code-base but start afresh and redesign how internal communication worked, particularly between threads and components relating to real-time processing. It's a very complex topic. If communication internals aren't clear, concise, and well-defined then bugs creep in. But during a live performance, a software crash is absolutely catastrophic: If you are on stage and your software crashes, the show is over!

ST: So… do your apps crash often?

HvH: Fortunately, no. If I discover a crash bug, I drop everything I'm doing and fix it. I have never had a report of a use-time crash.

In fact, if a crash-causing bug did exist, I wouldn't be doing this interview. I'd be fixing it! Live performance software isn't useful unless it's 100% stable, which is why reliability is a core principle of OpenAV software.

ST: What kind of music are you making with Luppp?

HvH: I'm into a lot of high-energy, fast-paced music: drum 'n' bass, dubstep, electro. I also play classical piano and cello, so the variety of music that I listen to is huge! Right now, I'm interested in the fusion of guitar + electronic music.

ST: Do you do live performances?

HvH: I've performed in the past, although not a great deal yet! I'm currently writing songs to perform at gigs and improving my confidence in front of larger audiences.

ST: What kind of music are other people making with Luppp?

HvH: Luppp is used by musicians of all genres, from ambient droning music to rock and guitar performances. One band is using Luppp for looping cello live, while the other member plays synths and triggers drum tracks. In short, there is no limit to what you can do with Luppp, so long as its looping.

ST: How do you fund your work on OpenAV Productions?

HvH: OpenAV Productions is now 18 months old. It was first announced at the Linux Audio Conference 2013. I've thought about how I could sustainably fund OpenAV development. At the moment, I'm writing the software to a feature complete 1.0 release – that means everything I think is essential will be finished. I have beta testers, and when they've ensured that the software is stable, I announce the existence of the project. I make promo materials (video, some screencasts, graphics that explain its features). I then request financial contributions in order to release the source code for the app.

I use a variation of the "bounty for source" model, but I include a time element as well. So, when the app is ready, I make a commitment to the community that one year later I will release it under the GPL, regardless of the amount of money that the project receives. That means I'm going to give you the app whatever happens. But if people give me money, I'm going to release it sooner, and everyone can use it sooner. So OpenAV doesn't sell software, and it doesn't sell source code either. What it actually sells, if that's the appropriate word, is time before the release. The motivation for people to contribute is getting the app sooner. It's not about whether they'll ever be able to use it, or ever get the source code; it's going to be released as open source anyway.

ST: What are the advantages of that funding model over more traditional ones, like one-off donations or crowdfunding, for example?

HvH: There are two key differences. One is that with traditional models, you're actually paying for the source code. The other is that with most of these Kickstarter or Indigogo campaigns, the product hasn't been created when people are asked to pay. You rely on the project's promises: "I can do this," and we hope that they're right. A lot of crowdfunded projects work very well. But the difference with the OpenAV model is that I've already built the app, it's finished and tested. The same day the target funding is reached, I roll the release, and it's available for everybody.

The Linux audio community is a very tight knit group of people. When my apps are ready, generally within a day or two, there is a binary available. This allows users to install it with a single terminal command or GUI package manager just a day after the code is released. By the way: huge thanks to all the packagers!

ST: Those are two benefits to your users. Are there any benefits of this model to you as developer?

HvH: Yes, I like to work with it, and prefer it to a proprietary licensing model for the obvious reason that there's an openness and transparency to how I operate, and a connection between my users and myself the developer. We're working together to make a tool that is useful to both of us. I don't have to try and market things to users who don't want to pay for them. When someone reports a bug I don't have to charge them a fee to upgrade to the fixed version. So, there's an honesty to this model that, in my opinion, is the right way to go. This is a creative way of working together on projects, and if people financially support my work then I'm delighted, and in return try to accommodate their requests. I create some demo content as well, to give people presets for the apps and helpful examples to work with.

ST: Do you seek to work full time on your audio apps? Would your funding model support that?

HvH: At the moment, the OpenAV release system is taking time to gather momentum. It needs to be something users are familiar and comfortable with. When I first announced the Sorcer wavetable synthesizer last year and laid out how the release system works, I received my first 10 Euros within 20 seconds of the announcement. That put a smile on my face – it was brilliant to see members of the community being so responsive!

But from what I've learned since then, do I see myself working on OpenAV full time in the future? Yes. It's a respect, trust, and a transparency thing, and it requires constant nurturing: something that I intend to continue doing for the foreseeable future. Will the OpenAV release system support full-time work? That all depends on trust, reputation, and respect between developer and user. This is built over time.

ST: Is that why the funding targets for your apps have been low so far?

HvH: Yes. For the first plugin I requested 120 Euros in donations in order to release it as open source. That involved 12 people paying 10 Euros each. The alternative would have been waiting a year. As it happens, within nine days I had 12 donations of 10 Euros, so it got released after nine days, which to my eyes was a fantastic success.

My time-release method provides users with an extra motive to contribute quickly, as opposed to waiting weeks or months after the announcement, and also to contribute more than they might have otherwise. Ten Euros per user is the amount I've asked for so far. But each 10-Euro donation also scraps one month of the delay to the release. The user pays a significant amount of money, but a significant delay is also avoided, so they feel that it's been worthwhile. That's a key part of this model: that users feel good after financially supporting our work.

ST: Because they're getting something valuable in return?

HvH: Yes. And, although only a handful of people financially supported the project, the rest of the community got access to the software, too! So when a user donates, they're actually supporting the release of a project for themselves and everyone else.

Rather than buying software in a conventional sense, they're supporting the developers, so the business relationship is reversed. You're not really "buying" software and receiving a copy in exchange for money. Instead, it's a healthier relationship where you're supporting ongoing work by sending money back to the project out of gratitude. It's a mutually beneficial situation.

ST: How do your earnings from this model compare to what you might earn as a professional software developer?

HvH: As I develop, I actually keep track of the amount of time that I spend behind the laptop. It's interesting for me to see how much money I have raised in total with OpenAV compared to how much time I've put into it. I've spent more or less 1,290 hours developing OpenAV software since the Linux Audio Conference 2013. That's actual coding time when I'm sitting behind the laptop.

In total, I've received roughly 1,120 euros (about US$ 1,400) in donations through OpenAV. So if you do the math, I've earned about 1.15 euros (about US$ 1.50) per hour for my programming work.

On the other hand, there is the issue that I mentioned earlier about building respect and trust from the community, and building up sufficient transparency so that people are comfortable in financially supporting the OpenAV project. We're not yet at a stage where the quantities of money I'm requesting allow me to work full time on this software. However, this is something that can change, assuming that users are comfortable with paying more for my work.

ST: Will the figures increase in the future to more accurately reflect the cost of development?

HvH: Yes, that's certainly something that I've been considering. It'll be a gradual process. I'm trying to build up donations to a point where a more sustainable and long-term funding model is possible. I'm passionate about that – brilliant communication between users and developers. We should all be working together to make awesome music software. That's the ultimate goal of the OpenAV release system: to create a two-way mutual benefit for both parties to cooperate and to work together.

On the support page [1], users can donate money at any time, which helps OpenAV become financially viable.

ST: Let's move on to Linux audio as a whole: You've spent a long time in "the scene."

HvH: 2009 was my first Linux Audio Conference. I haven't missed one since. It's a brilliant resource, a brilliant conference, a brilliant time to meet everybody, to see new things, to bounce ideas off each other, to ignite new collaboration, to find inspiration, motivation; it's a fantastic time and I really, really enjoy being part of it.

ST: Are you happy with how audio apps have progressed since 2009?

HvH: Yes, totally! At the time, I was starting out in the Linux audio ecosphere, I'd done maybe a year or two of C++ audio programming, and the conference was the first time I'd seen the famous guys of Linux audio, you know, the guys who built the frameworks that everyone else integrates with. There's been huge improvements since 2009, in fact, too many to mention!

There has also been a huge increase in the number of synths and plugins since then, more development in general, and the community is in an even better place. A lot of people are very actively working together. There's a lot of communication. I don't think that a single day goes by without useful discussion on IRC (By the way, I'm harryhaaren in #lad on irc.freenode.net ).

There are many places where you can see the progress happening before your eyes if you know where to look, and I've got high hopes for the future of Linux audio. There are a lot of new ideas and concepts for workflow within the scene that could spring into really interesting new ways of performing and making music.

ST: Do you think that the number of Linux audio users is growing as well?

HvH: We regularly see new names popping up on the IRC channel, mailing lists, and personal email. I get emails from users from time to time saying they've just discovered my software, asking me to port it to Windows and Mac, or "I'd like a plugin that does this." Interest is certainly growing.

ST: What's most exciting for you about Linux audio right now?

HvH: There's almost too many projects to talk about! There's the MOD project [2], a fantastic project to build a hardware guitar effects pedal which is very modular and scalable, based totally on open source frameworks and running Linux on the inside. The MOD project has recently been successfully funded on Kickstarter, which means that very soon ArtyFX will be taking to the stage in the form of a stompbox!

ST: What else?

HvH: The Faust language is getting more powerful by the day. There's more integration between projects, like LV2 plugin specification, various DAW hosts. There are ongoing discussions about how we can best tighten the coherence and user experience between all these projects to make an absolutely awesome way of producing music, for the users and for ourselves.

Everyone's working on this for a common goal. And that's something that, especially during this year's LAC even more than during others, you can feel that there's that buzz of a community that are trying to achieve new things that haven't been done before.

ST: Do you view Linux audio apps as competition for non-free apps?

HvH: Not really. I see tools as a choice. It's like having two brands of any product – they serve the same purpose, but one may be slightly better for your particular needs. What I'd like to see is projects working together to provide different workflows for users who have different needs. Rather than competing with non-free apps, I'd be interested in working together.

At the end of the day, I want to make music. I hold no grudges against people who use non-free software to make theirs, and I understand that the world of Linux audio is not the most suitable platform for everyone right now. Those who aren't accommodated are the one's I'm most interested in actually, because I can learn most from them. It would be fantastic to get more communication going with those people.

ST: So, Linux audio should just "be better" in order to attract more users?

HvH: Do we need to try and "beat" the Mac and Windows platforms? No, that's not what it's about, not at all. It's about making music. The tools that allow you to do that are the best ones. Period. I know that does go against a Free Software principle that I personally find important: that I can modify the code myself, that I can see how it works. But everyone has their own workflow, and their own way of doing things, and of course I respect that, too. I would encourage users who haven't tried Linux audio yet to grab a copy of Ubuntu Studio [3], for example, and try it out!

ST: What's next for OpenAV?

HvH: OpenAV is currently working on a new wavetable synthesizer called Ninja [4], with anything to anywhere routing. It's very flexible, with a lot of wavetables going in there, so it should be general purpose and very useful to a wide variety of people and styles.

Apart from that, OpenAV is always experimenting with new features and ideas, workflows, and musicians. Right now, we're working on a one-man-band solution which will hopefully change th0e way solo performers work on Ubuntu. There are live-visual and lighting projects in the pipeline, as well as feature updates and bug fixes to the projects we've already released.

Figure 1: Harry van Haaren demos the OpenAV music suite.

Figure 2: Harry van Haaren introduces the OpenAV suite at the 2014 Linux Audio Conference.

Figure 3: The delegates, musicians and developers at the 2014 Linux Audio Conference.