This is the project website & blog of MIRLCAuto: A Virtual Agent for Music Information Retrieval in Live Coding, a project funded by the EPSRC HDI Network Plus Grant - Art, Music, and Culture theme.

Presence at WAC 2021 – July 5-7, 2021

09 Sep 2021 • Anna Xambó • Post a comment

MIRLCa on dice. Screenshot from one of the two performances at the Web Audio Conference 2021, Barcelona, Spain, via Zoom.

This blog post presents the project's presence at the Web Audio Conference (WAC) 2021, held in Barcelona, July 5-7, 2021.

Keynote #

This year I had the honour to be invited to give a keynote talk at the Web Audio Conference 2021, which was programmed on July 5, 2021. The keynote is entitled "Grassrouters Music Performance" and reflects on my previous work on network music, audience engagement, and live coding.

Due to the still ongoing pandemic, the conference was online and the WAC 2021 committee suggested adding interactive components during the session. I designed the presentation on Miro and prepared 3 small capsules with a participatory element at the end of each of the three main parts of the talk. You can watch the presentation on the following video.

The other keynote speakers were Anna Huang, who is Research Scientist at Google Brain, and Juanjo Montiel, who is a Senior Software Engineer and Microsoft Accessibility MVP. You can find the full information here.

Performance #

I had also the opportunity to perform a live coding session with MIRLCa at WAC 2021. The challenge here was to design a performance to be delivered on Zoom. In order to cover the full global timezone, the performance was scheduled twice on the same day, on July 6, 2021. I decided to perform live on two occasions.

Performance 1 #

Performance 2 #

In the two performances, I explore chance by using a set of four dice that trigger a number used to retrieve sounds from Freesound by its sound identifier. The paradox, in this case, is that the agent that filters sound by preference does not have sense with this command. In the follow-up discussion of the first performance, Frederic Font from suggested that you could still retrieve the best candidate if the current target was predicted as a bad sound. Something to consider for the future, certainly.

Here you can find direct links to the keynote slides and the performance paper:

Acknowledgements #

Thank you very much to the WAC 2021 committee, who successfully organised and delivered an online conference exploring the possibilities that the online medium can offer. Special thanks to António Ramires for hosting the keynote session, Ángel Faraldo for hosting the two performance sessions, and Frederic Font for his constant technical support with Freesound. Also thank you to all the WAC 2021 attendees who participated in the interactive capsules of the keynote talk as well as attendees of the one or two performances. Last but not least, thanks to Gerard Roma for his valuable feedback.

← Home