Web Audio Conference (WAC) 2017

Aug 30, 2017Audio Technology, Blog

In today’s digital age, searching for new technologies is easier than ever. The web is flooded with sites (Reddit, Twitter, StackOverflow…) that offer answers to questions you did not even realize you had. However, no source of information on the web can be compared to meeting and discussing with your peers in person.

That is exactly what the Audionamix team did August 21st – 23rd at the Web Audio Conference (WAC) 2017. The third annual WAC was hosted by the Queen Mary University of London: an engaging location to listen to presentations, meet people and have a couple of drinks.

WAC at the Queen Mary of London

The conference brought together people of different backgrounds (researchers, developers, artists), but all interested in one particular technology: the Web Audio API. This tool, embedded in web browsers, is relatively new as its first working draft was made available in late December 2011. With lots of room to grow, its community is dynamic and eager to help the API evolve.

We had two goals for the conference: to evaluate the current Web Audio Technology, and to discover tools that can help our everyday work here at Audionamix.

Is Web Audio Ready for Professional Audio?

When evaluating a new technology, a big requirement is whether it is ready for professional use. To this end, we attended three presentations of advanced tools to the WAC community.

Michel Buffa (Université Côte d’Azur, CNRS, INRIA) presented his Real Time Guitar Amplifier simulator. Based on the Marshall JMC 800 amplifier schematics, it behaves just like classic simulators. It was even used in real gigs to test its stability. It is a perfect example of advanced developments using the Web Audio API. More details can be found on his publication here.

Next up was the SoundTrap team. They presented their in-browser Digital Audio Workstation (DAW), another impressive example of advanced development with Web Audio. As a browser-based software, SoundTrap offers collaborative features that would be very difficult to integrate in standard offline, local DAWs. As a popular and attractive tool, especially in the educational community, SoundTrap is a perfect example of successful software based on the Web Audio API. More information is available on their website.

Finally, Matt McKegg presented Loop Drop – software that combines a looper, synthesizer and sampler and is designed for live performances. In his presentation titled “Is Web Audio Ready for the Stage/ Dance Floor?”, Matt presented the limitations he experienced in coding his software. He explained that, as JavaScript features a garbage collector, you cannot predict how much CPU your tool is going to use at a given instant. Because of Murphy’s law, the garbage collector is likely going to start at the wrong time and produce glitches in your audio output. More information about the tool can be found on his website.

François Observes the Loop Drop presentation at WAC 2017.

François observes the Loop Drop presentation at WAC 2017.

Following these presentations, we do not believe that the Web Audio API is suited for professional applications quite yet. However, considering the speed at which the API is improving and its attractiveness to high-caliber developers, we are confident it will be at some point in the near future. As Paul Adenot from Mozilla noted in his presentation “Recent and Future Evolution of the Web Audio API, their API developers are well aware of its limitations and are working to address them.

How Can Web Audio Help Us?

Although the Web Audio API is not ready for professional use, that does not mean it cannot or should not be used at all. It is easy to see how the API can be a huge help for simple tasks, as tools like trackswitch.js demonstrate. Presented by the International Audio Laboratories of Erlanger, trackswitch.js is a versatile web-based audio player for presenting scientific results. The community is also constantly building more and more open source tools to make the API even simpler and more powerful.

The team listens to presentations on trackswitch.js

The team listens to presentations on trackswitch.js.

On the research side, the presentation made by Michel Buffa (Université Côte d’Azur) about the two million song database project called WASABI was also very exciting. Now that machine learning is at the center of many technologies (ADX Technology included), this kind of database that would integrate features from many songs has great potential for training algorithms.

Overall, the trip across the Channel to the Web Audio Conference was educational and fruitful, both in learning the limitations and potentials for the Web Audio API and Audionamix.