Thuistezien 249 — 28.04.2021
Instrumental Shifts
Agnusdei & Dideriksen
Agnusdei & Dideriksen
The 2019 ‘Instrumental Shifts Symposium’ – a collaboration between RE:VIVE, Rewire and West Den Haag – was the first academic symposium to be featured in the Rewire festival’s programming. With Rewire’s dedication to uncovering exciting musical acts and new adventurous creative directions, the festival attendees enjoy performances which frequently demonstrate cutting edge electronics. This made it only natural for the festival to invite interested audience members even deeper into the complex and intricate world of contemporary electronic music by opening the floor to selected researchers, music scientists and artists to discuss their work on developing new music technologies. One of these talks was given by professor Bob L. Sturm of Stockholm’s Royal Institute of Technology, during which he presented his work on the ‘Folk RNN’ project. His talk was accompanied by a performance which responded to the aforementioned project, by two artists who at the time were both based in Den Haag: Italian saxophonist, composer, and electronic musician Laura Agnusdei, and Danish composer, conductor and fiddler, Lauge Dideriksen.
Folk RNN is a freely accessible website that allows users to generate music using a recurrent neural network (RNN). In this application, this form of artificial intelligence digests transcriptions of folk music to then generate compositions in a stylistically similar vein. Two databases of musical transcriptions are used for this, one of which primarily holds transcriptions of Celtic folk music, and the other mainly Swedish folk music. The output is affected by decisions made by the user, such as the choice of which database to use, but ultimately the program outputs a melody. If interpreted in an idiomatic folk music style, this can become a fully-fledged composition that can often convince listeners that they are genuine relics of a folk music’s tradition. In examining the success of the system, Professor Sturm found that many folk music enthusiasts would enjoy the compositions they would hear, as long as they weren’t told that the melodies forming the basis of the compositions were generated by a machine.
Professor Sturm himself is an accordion player who has shared performances of himself playing Folk RNN tunes in an Irish folk-minded manner. His performances can easily cause the listener to forget that any machines were part of the process whatsoever. However, Agnusdei and Dideriksen take a different approach to engaging with material produced by Folk RNN in their performance, by exploring the space between the digital and physical realms of the music that the system is inherently a part of. In the hands of these two artists, machine generated folk melodies move between the realm of acoustic, idiomatic-feeling folk, and a world of abstract electronic sound textures and live sound processing. Dideriksen himself is a highly skilled performer of traditional European dance music, and although he starts the performance by creating abstract screeching tones on a violin that one might typically expect to see in a classical concert, he later moves to the Hardanger Fiddle (a traditional violin-like instrument used originally to play the music of Norway), and draws us into a world of Scandinavian-feeling folk sounds. On the other hand, Agnusdei juxtaposes and compliments this with electronically manipulated versions of the live fiddle sounds that spread out into reverberant sonic textures or chop the fiddle audio into chaotic looping snippets of sound, as well as other electronic textures that she generates with material from Folk RNN – her methodology is also one that can be seen in her solo performances, where her transparent and melodious saxophone playing alludes to hints of stripped-down chill jazz, exotic-feeling folk and even dub and reggae, and act as a foil to the complex and slow-moving electronic tapestries that she unfurls through manipulating knobs and buttons live on stage.
Whereas Professor Sturm is tempted to hide the machine system in his performances, Agnusdei and Dideriksen thrive on this hybrid space of digital and traditional-feeling. They use it to move us into a mysterious, strange and beautiful digital landscape that engulfs, complements and alters echoes of folk-tinged melodies, as if they are wandering through the subconscious of our 21st century hybrid electronic culture.
Text: James Alexandropoulos - McEwan
Folk RNN is a freely accessible website that allows users to generate music using a recurrent neural network (RNN). In this application, this form of artificial intelligence digests transcriptions of folk music to then generate compositions in a stylistically similar vein. Two databases of musical transcriptions are used for this, one of which primarily holds transcriptions of Celtic folk music, and the other mainly Swedish folk music. The output is affected by decisions made by the user, such as the choice of which database to use, but ultimately the program outputs a melody. If interpreted in an idiomatic folk music style, this can become a fully-fledged composition that can often convince listeners that they are genuine relics of a folk music’s tradition. In examining the success of the system, Professor Sturm found that many folk music enthusiasts would enjoy the compositions they would hear, as long as they weren’t told that the melodies forming the basis of the compositions were generated by a machine.
Professor Sturm himself is an accordion player who has shared performances of himself playing Folk RNN tunes in an Irish folk-minded manner. His performances can easily cause the listener to forget that any machines were part of the process whatsoever. However, Agnusdei and Dideriksen take a different approach to engaging with material produced by Folk RNN in their performance, by exploring the space between the digital and physical realms of the music that the system is inherently a part of. In the hands of these two artists, machine generated folk melodies move between the realm of acoustic, idiomatic-feeling folk, and a world of abstract electronic sound textures and live sound processing. Dideriksen himself is a highly skilled performer of traditional European dance music, and although he starts the performance by creating abstract screeching tones on a violin that one might typically expect to see in a classical concert, he later moves to the Hardanger Fiddle (a traditional violin-like instrument used originally to play the music of Norway), and draws us into a world of Scandinavian-feeling folk sounds. On the other hand, Agnusdei juxtaposes and compliments this with electronically manipulated versions of the live fiddle sounds that spread out into reverberant sonic textures or chop the fiddle audio into chaotic looping snippets of sound, as well as other electronic textures that she generates with material from Folk RNN – her methodology is also one that can be seen in her solo performances, where her transparent and melodious saxophone playing alludes to hints of stripped-down chill jazz, exotic-feeling folk and even dub and reggae, and act as a foil to the complex and slow-moving electronic tapestries that she unfurls through manipulating knobs and buttons live on stage.
Whereas Professor Sturm is tempted to hide the machine system in his performances, Agnusdei and Dideriksen thrive on this hybrid space of digital and traditional-feeling. They use it to move us into a mysterious, strange and beautiful digital landscape that engulfs, complements and alters echoes of folk-tinged melodies, as if they are wandering through the subconscious of our 21st century hybrid electronic culture.
Text: James Alexandropoulos - McEwan