The future of Ardour

Ardour is likely the most compelling open-source digital audio workstation (DAW) for music professionals. But a recent blog post by Ardour's lead developer, Paul Davis, revealed that he will likely need to shift his focus due to a lack of financial support for the project:

Given the support from users and companies for the project, that news comes as a bit of a shock. Users have stated that Ardour can hold its own against Pro Tools, a proprietary DAW used by professionals throughout the music industry. It has also been used in the Mixbus DAW product from Harrison Audio Consoles, sales of which provides some income for the project.

Davis is not completely stopping development work on Ardour; he has "the option of working for a digital audio company that is developing new projects based on Ardour". However, this would focus development on the needs of that particular business over Ardour's end users:

Davis concluded the post with uncertainty about the prospects of the community picking up the burden, and an insistence that his message is not a request for funding. He also emphasized in a comment on the post that he is not abdicating his role as lead developer. However, the message is clear: Ardour's lead developer will likely shift gears on that work in the near future.

What happened? How did a project that received both commercial attention and high praise from users not find the means to fund one developer full-time? To help figure that out, we will need to take a look Ardour's history.

Davis had started the project in 2000, working full-time on Ardour for several years, buoyed by a windfall from his work for Amazon.com in its early days. But the issue of financial sustainability would rear its head before Ardour's initial "stable" 0.99 release in September 2005 (developers made a decision to skip the 1.0 release and make major changes for a 2.0 release). In a post to the development mailing list in May 2005, Davis explained that he had only earned $6,000 over the past five years to support his work. As a result, for a time, he would have to take an unrelated development job that would consume most of his time during the week.

In 2007, the project started a subscription-based funding model. Web site visitors could only download binaries with a paid subscription, which currently costs $1, $4, or $10 per month, with a $50/month option for institutions. Full source code remains available for free download. However, with the project only targeting Linux and Mac OS X users, this seems not to have led to a sustainable model, as popular Linux distributions packaged Ardour for their users. For example, in an October 2007 LWN review of Ardour 2, Forrest Cook reported that the multimedia-focused Ubuntu Studio came with Ardour out-of-the-box.

The community remained concerned about sustainable funding, with the topic dominating discussion on the development mailing list in January 2009. Patrick Shirkey of Boost Hardware then suggested a number of possible funding sources, ranging from seeking grants, to a music CD featuring artists who use Ardour, to celebrity endorsements.

While Davis's income from Ardour had improved by June 2009, an interview with Linux Journal around that time revealed that he was still very concerned about his personal financial situation. Positive attention toward Ardour, coupled with concern for its self-sustainability, remained during the years to come. In a 2010 episode of Jono Bacon and Stuart Langridge's "Shot of Jaq" podcast [.ogg], they noted that software projects relying on funding directly from end users to finance long-term development have struggled in comparison to the proven model of corporations financing free software projects while selling related services (e.g. Red Hat and SUSE). There have also been marketing issues with the subscription model, with one potential user complaining in 2011 about having to pay for binaries for this particular open-source project when other open-source software is available free in both source and executable forms.

In 2012, we finally saw an Ardour-based release for Windows users: a proprietary, closed-source product named Harrison Mixbus. Windows users arriving at the Ardour download page were directed to the Harrison Consoles web site to purchase the product: it currently costs $149 outright (discounted from $219), or $49 plus $9/month for a subscription.

There was also discussion in 2012 about charging $10-20 for Ardour on the Ubuntu Software Center, which did not lead to doing so; discussion was short-lived and the idea did not seem to be taken seriously by the community. Ardour's 3.0 release came in 2013, with many new features, such as complete support for MIDI.

It appears that Ardour's subscription model has one major technical flaw, which may have cost the project some money. Some formerly subscribing Ardour users, who were concerned about the future of the project,commented on Davis' recent post noting that they had various technical difficulties with the subscription mechanism. Davis replied that PayPal, which is used for subscriptions, does not provide a programmatic interface to view canceled subscriptions, leaving him to manually update the subscription database by downloading a CSV file from PayPal. Addressing that issue may pull in some extra cash, but likely not enough to cause Davis to reconsider his decision to take the new job.

Perhaps an opportunity was missed with the project's refusal to consider "average" Windows users (i.e. those who couldn't or wouldn't pay the high cost for Harrison Mixbus) as a potential userbase worth targeting. In 2006, Davis noted, albeit without providing examples, that several open source projects offering Windows ports have not had the capacity to manage the increased demands:

In 2009, Davis again argued that the social costs of supporting a large, non-technical Windows userbase who are not as "willing to support software developers who provide useful tools" outweigh any benefits to the project. But with the massive install base of Windows, the standard subscription option might have captured quite a few Windows users. It's plausible that, with a marketing push, Ardour could have found enough Windows users to subscribe at, say, $10/month. That might have brought in enough to enable Davis to stay on full-time. Or it may not have; that influx might have been more than eaten up by the additional support burden, which is what concerned Davis.

In 2011, Davis noted that the market for DAWs is small, and that even the mainstream proprietary DAWs have been financially struggling:

It is sad to see such a well-respected open-source project — one that is key to attracting certain users to Linux (i.e. professional and hobbyist musicians) — be unable to support even one full-time developer. The history of Ardour serves, in part, as a warning to potential developers for open-source software niche projects; relying on end users for funding paid work may not be a sustainable model.

Note: you can avoid this step in the future by logging into your LWN account.