Audio Compost a collaborative virtual frippertronic loop

Audio Compost a collaborative virtual frippertronic loop August Black Department of Critical Media Practices University of Colorado, Boulder [email protected]lack ABSTRACT Audio Compost is a web application that enables participants in a single location to contribute sound to an ongoing, collaborative frippertronic loop using their mobile devices and voice. The system tightly synchronizes audio samples across devices using a web object timing server and operates in two modes, centralized and distributed. In the centralized mode, participants send sound to one main audio output and their mobile phones are silent. In the decentralized mode, participants use their mobile devices for both input and output, forming an impromptu swarmed multi-channel audio system. The resulting auditory landscape of voices, noises, and ruminations varies based on the number and composition of participants as well as with parameters set in the system. This paper explores the development, implementation, and performances of Audio Compost at open-air venues and planetariums, offering insights into its interactive and dynamic sonic environment. 1. INTRODUCTION Audio Compost is a software and network service that creates a virtual repeating and diminishing loop among participants. Partly inspired by the context in which this instrument was conceived - the 2022 Digital Naturalism Conference - I call it ”Audio Compost” because I see it as a way to convert stored or unrealized acoustic energy into decomposed ”organic fertilizer” for the sonic imagination. With this web-based software built on Web Audio technology, participants in one on-site location - but from potentially different backgrounds, identities, and musical inclinations - speak or sound into the portable microphones on their mobile phones to feed the system with acoustic content of their own choosing and making. The system mixes the incoming sonic material from the network into an ongoing frippertronic looping process that is tightly synchronized across all devices using the Web Timing Object API[10, 11]. Depending on the size and makeup of the audienceparticipants, the sonic result is sometimes rhythmic, sometimes cacophony, and sometimes awkward, until it is not. As the audience learns to improvize with the system in an ongoLicensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0). Attribution: owner/author(s). Web Audio Conference WAC-2025, November 19–21, 2025, Paris, France. © 2025 Copyright held by the owner/author(s). ing act of discovery, an assemblage of hoots, howls, whistles, stomps, bomps, and swoons unfold and fade ad infinitum into and out of recognizable patterns or structures. It is as much a system of social connectivity as it is acoustical in nature. A video documentation of the last performance in 2024 can be found at https://assets.august.black/media/ compost/wyoming edit2024.mp4. Other loop-based collaborative or mobile work exists as prior art for this research. [29, 20, 19, 39, 40, 31, 17, 13] Audio Compost presents research that combines voice, timing, and transport in new ways along with design, scalability, and format to enact performances in open-field and planetarium contexts that can potentially involve 100’s of users at once. In what follows, I describe the motivation and implementation of the work. I also discuss the results and innovations while articulating some insight gathered from various performances with the system. 2. RELATED WORKS Audio Compost is influenced by a range of open-ended participatory systems, from Fluxus happenings to contemporary mobile applications that work with voice, samples, or live streams. The compositional approach of overdubbing and looping has an openness that lends itself well to many of these practices. Below, I review a specific kind of looping technique that informed this work, called Frippertronics, and then discuss more generic participatory systems in the audio domain that I call “connected musics” before discussing contemporary participatory systems with mobile devices. 2.1 Frippertronics Frippertronics is a tape looping technique, named by English guitarist Robert Fripp[4] and ambient musician Brian Eno in the 1970’s, but originating earlier in the mid-20th century. Tony Conrad used the technique in 1961 for multiple compositions titled “Three Loops for Performers and Tape Recorders” and Pauline Oliveros used the it for “I of IV”in 1966. In 1963, Terry Riley called the technique “Time Lag Accumulator” in his music for Ken Dewey’s play “The Gift”. Eno introduced the method to Fripp when using it as a means to produce slowly developing background layers of sound for Fripp’s guitar playing on their 1973 duo album ’No Pussyfooting’. (see [37]) The technique uses two reel-to-reel tape decks, with the record head of the first deck connected to the playback head Figure 1: Diagram of the tape recorder set-up from Tony Conrad’s score of ’Three Loops for Performers and Tape Recorders (1961). of the second in a physical tape loop. See Figure 1 for an illustration. It is a simple and versatile acoustic system that allows for the ongoing collection of arbitrary sonic impulses in the aggregate. Unlike traditional looping in dance music or standard delayed feedback, these impulses acquire compositional structure over time that are muddy, blended, and faded. It can be used to build layers of sound from one or more instruments or acoustic devices that repeat and decay over time and lends itself well to various flavors of what one might call the ambient genre. My first encounter with an analogue frippertronic system was at Documenta X in 1997 in Kassel. An artist by the name of Sergio Messina produced a live interactive audio work with the ORF Kunstradio[3] and Radio TNC where he spliced a long analogue tape over two recorders and enticed the public to speak into an open microphone. The setup had a physical presence in the room that was partially awkward and intimidating, but also alluring. The simplicity, versatility, and liveness of the technique is the key to its appeal. The results as people spoke into the system were rhythmic and playful and kind of a natural setup for audience interaction that connected voices and intentions in one setting. It wasn’t always musical per se, but would acquire structure on the periphery of what one might accept as musical. This desire to connect voices or music making capacities together has long musical tradition, but may be most apparent in something I call “connected musics”. 2.2 Connected Musics In the section “Ears and Clairaudience” from R. Murray Schaffer’s introduction to his book, Tuning of the World, he says “hearing is a way of touching at a distance and the intimacy of the first sense is fused with sociability”.[36] It is the unbounded, undetermined and ambient nature of sound that allows people to use it to connect with one another. This is almost always a social phenomena, even when transmitted through networks. There are a number of Avantgarde participatory networked music practices of the same lineage that inform this research, and that start with openended acoustic systems and continue with more technologically driven works. This research finds inspiration from the playable devices by “participants” in John Cage’s Imaginary Landscape no. 4 (1951), Laurie Anderson’s Car Horn Symphony (1969), Jose Maceda’s Ugnayan (1974), site-specific sound projection and networking pieces by Bill Fontana, early radio work by Max Neuhaus, as well as telecommunication work of Robert Adrian X as prior art for distributed and or participatory systems. Some of this history is discussed in [43, 14, 12, 21, 5]. I am particularly inspired by early telematic and radio work in the 1990’s by Heidi Grundmann’s Kunstradio, especially the organization of Zeitgleich, Horizontal Radio, and Rivers and Bridges.[49, 3, 22] It is in projects such as these that one can understand the migratory and cross-border aspects of auditory landscapes in the broader scales of the electronically connected world. The transitory and transmissive nature of sound in motion gives it a sense of connectedness and a sense that it has stored potential to be everywhere all at once, albeit physically constrained. In networked environments, musical energy is found not only where it is supposedly intended, but also in the interstitial places where sounds overlap and collide. Audio Compost is an attempt to find new kinds of intersticial sound “spaces”. 2.3 Mobile Device Participation and Output More concretely, prior research using mobile devices for collaborative musical events, even before the smartphone era, informs my work. Some examples are Golan Levine’s Dialtones symphony that used the audience ring tone as musical content, [26], Atau Tanaka’s early hacked mobile phone engine[42], interactive music software on mobile phones[38], the usage of in-phone microphones[8] early mobile GPS [46], creative sensor mapping on mobile[18] mobile phone orchestras[28], etc. This early work investigated the new mobility of sound recorders and playback devices. With the proliferation of mobile smartphones around 2007-2010 along with the invention of Web Sockets in 2011 that provided bidirectional communication in the web (outside of early macromedia flash), more interesting research explored dynamic real-time participation. The mixture of available networks with pocket-sized touch-enabled mobile computation allowed for everything from collective instrumentation [45, 24], to percussive drumming [25] to innovative mobile UI design [33, 32], spatial proximity [7], ensembles [9, 35, 16], game play [15], immersive environments for distributed performance [34], distributed music instruments [47], and general reflections on composing with audiences[48], among other things. There are too many to cite in this short paper. Particularly relevant work are various mobile based or collaborative loopers such ARLooper, LoopBoxes, LoopJam, Collective Loop, and Living Looper. ARLooper can visualize recorded sound as 3D waveforms in an AR space, but is unfortunately Apple-specific, only working in lighted areas, and untested at scale or in performance situations.[29] LoopBoxes is an accessible digital musical instrument designed to create an intuitive access to loop based music making for children with special educational needs[20]. My hope is to have Audio Compost work as fluidly and instinctively as this research. LoopJam allows participants to interact with a sound map using a 3D computer vision tracking system.[19] Collective Loop presents a collaborative co-located set of applications that invite users to engage with audiovisual media using mobile devices.[39] Finally, Living Looper is a real-time software system for prediction and continuation of audio signals in the format of a looping pedal.[41] Other relevant work forms sound from participant devices in various ways. One of my favorite participatory system of mobile devices is Fields[40] that focuses on sonic diffusions over mixed devices. Two feedback based collaborative mobile works are GroupLoop and Squidback[31, 17] GroupLoop is a browser-based, collaborative audio feedback control system for musical performance that builds modifiable feedback loops between WebRTC participants in a mesh network. Squidback as another mobile feedback work that creates live ambient music with open mics from participant devices and user-modified spectral suppression.[17] Recent research that lives online as a virtual memorial commemorating victims of mass shootings is also relevant for the kind of togetherness it creates through auditory landscapes as sites of reflection.[44] This work uses audience devices as a multi-channel system where forming a group is required to listen to the sonic memorial in total, and as means of remembering with others. The audience in these mobile systems is more than just a speaker array. They are, or can be within certain configurations, lively actuators with agency. In these examples, the sonic landscape exists where and when the people are connected and sounding. It is not just about the sound in sedentary situations, but the social connection that activates and shapes a moment of time. While these applications above explored looped sonic event structures in multiple ways, sometimes as installations or even over distributed over mobile phones like my prototype, Audio Compost takes a divergent approach and is unique in that it provides a particular kind of live mediated looped musical experience with open mics that one can perform on-the-go using only mobile phones and at scale (given proper networking) with or without some sort of conductor leading the exchange. I also believe the interface and network topology has a simple but innovative form-factor and design that is different from prior art. 3. PROJECT HISTORY AND DESCRIPTION I first prototyped the distributed mode and performed with Audio Compost in Batticaloa, Sri Lanka at the 2022 Digital Naturalism Conference (Dinacon)[30], a month-long event that brings together field biologists, interaction designers, engineers, artists with the purpose of innovating on jungle craft in various practical and creative dimensions. I had originally planned to work with others on a shoreline radio program using found sounds on radio frequencies and field recordings, but got sidelined by a covid quarantine. Working alone for the week, I put together a prototype of Audio Compost that created a three minute long virtual frippertronic system and allowed users to record their voice on their mobile devices, send the audio to a ’central’ edge server (cloudflare worker with R2 cloud based object storage) via simple HTTP posts, and where a timing mechanism based on the Web Timing Object tightly synchronized the output. 1 I built a simple interface where the left-to-right of the screen represented a 20 second loop, and an animated blue line represented the time moving forward moved from left to 1Cloudflare workers provide a globally geo-distributed architecture that allows exchange to happen at a close proximity to users. Figure 2: Initial Audio Compost interface with blue line indicating time, and yellow bars indicating audio segments. A yellow strip will turn green when playing. Figure 3: The first run of the system at Dinacon 2022 with lively participation and users instinctively forming an audio swarm for each other. right. I used a basic Web Audio setup with BufferSource and Gain nodes. A red ’record’ button overlaid on top allowed a user to press it and record sound (using the MediaRedorder API) until they pressed stop (or they recorded a max of 20 seconds). Each time the user recorded a segment, a yellow vertical bar would represent the start and stop times on the horizontal axis (see Figure 2). At the end of recording, the app would HTTP POST the audio to R2 and the cloudflare worker would tell any other client in the system via Web Socket to fetch the new audio via HTTP GET and put it in their respective timelines.(See Figure 7) The system was optimistic in that it never waited for all of the clients to download the new audio. The Web Timing Object in the code would keep all devices in sync using a Web Socket with a central server in Norway2. Since the user devices collectively project the sonic output in this distributed mode, any missing audio on a single device wouldn’t matter much to the sonic whole and would eventually download and fill in at some point. When an audio entered the system, the app would set it in the system and play it at full volume the first time around. Each time the app looped at its 20 second intervals, that 2Due to how the Web Timing API calculates and compensates for delay, the central timing server in Norway was never an issue Figure 4: Audio Compost mobile UI showing record, stop, and wait modes. same sample would be played at lower and lower volumes until it was inaudible and left the system, mimicking the degraded overlapping playback of frippertronics. When a sound played on any device, its representational yellow bar would turn green to indicate activity and its opacity would be set relative to its volume (opaque green at full volume and near transparent at no volume). Each sound would play a total of 9 times altogether for a total of three minutes in the system. The initial performance was well-received and lasted over 20 minutes with various waves of sonic content. Sometimes participants would make noises to mimic what they already said or what others said. Sometimes they would use words to answer each other in a call and response style. (See Figure 3) I was able to repeat this performance at the 2022 PIFcamp in Soˇca, Slovenia; again outdoors with a group of over 20 energetic and creative individual and with similar results[27]. Later in 2023, I was invited by Arte, Ciencia y Tecnologia y CK:\WEB de Idartes [2, 1] to adapt the system for a planetarium setting. I kept the main premise of the system - that sounds enter via mobile phone uploads into a 20 second loop, get played 6 times total (3 minutes) with decrementing volume before leaving. However, I modified both the interface and the internal transport heavily for centralized playback. For the UI, I changed the look and feel of the interface to better adapt for the darker ambience of a planetarium. I changed the background from mostly white to mostly black. Instead of yellow bars representing the sounds, I assigned each participant their own color. This not only added to the visual dynamics, but allowed each participant to follow their own recordings on the centralized screen of the dome.3 It also allowed them to disambiguate the audio of others. Additionally, instead of just rectangular shapes to represent the audio on the timeline, I showed a minimal waveform of the audio in the timeline. For the 360 dome projections (See Figures 6 & 9 ), I created a new central admin visualization. Essentially, it is just another client and can participate like the others. While each participant sees a rectangular interface on their phones where they contribute recordings, the dome client looped in 360 fashion like a radar system. The 20 seconds from left to right on the rectangular client mapped directly to the 20 seconds for the indicator line of dome visualization to circumnavigate. When shaped to the planetarium’s 360, I noticed the audio waveforms looked somewhat like clouds in a landscape, so I decided to slowly animate them from bottom of the planetarium screen to the top over the full two minutes they exist in the system, until they leave. This 3Thanks to Carol Sabbadini for the suggestion. Figure 5: Dome projected version from live performance at the Wyoming Planetarium. The pinkish line indicates time. As time moves forward the waveforms animate and move (float) towards the top of the dome, like thoughts offered to the heavens. effect in the planetarium kind of makes the visuals take on a religious feel, like voice clouds floating up to the gods and goddesses. Conversely, on the rectangular client, the waveforms animate slowly from bottom to top over the six 20 second cycles. Another addition I made was to the transport and tempo. In the prior two iterations, there was no limitation on how much one could input into the system. If a user finished recording they could immediately start a new recording. Without record limitations, and even with just 20 participants, the audio can build a little too fast if there are some eager contributors. There always are! For the new transport, I limited the interface to wait a random amount of time between 7 and 20 seconds from the time the user stops recording until they can record and input again. (See Figure 4) I also switched the uploading transport from HTTP to binary Web Socket so that I could draw the incoming audio’s waveform to the central visualization as it comes in. At the transport level, I did this using a portion of the MediaRecorder API that allows you to start with a ’timeslice’ variable and thereby sending batches of mono opus-encoded audio at one second clips at a time for low bitrate interaction[6]. Besides speeding things up, I was able to remove the R2 persistence layer and minimize the system complexity. (See Figure 8) Additionally, I can optionally reflect that audio back to the other clients for their displays for distributed mode playback in open-field situations. This optional ’bi-directional’ mode allows the system to perform better in low-bandwidth or large scale centralized situations like in the planetarium or in auditoriums. Figure 6: 2023 Performance at the Bogot´a Planetarium 4. DISCUSSION I’ve set this performance system up on a few different occasions and have learned from each. The first time I ran the initial prototype, I was within a conference that included an amazingly creative group of 20-30 individuals and their kids. Within seconds, everyone understood the gist of the ecosystem and were able to freely and energetically fill it with audio, almost to a fault where the sound become one massive drone with moments of audible chatter. Because there were no limitations on input and users were enthusiastically adding more and more, the system got overloaded. It was easy to hear the first round of the initial recording, but very hard to distinguish the same sound when it came back around again. There was also a learning curve involved with scaling the system up to handle larger audiences. At the scale of 100 or more participants much attention must be placed on the local wifi. Multiple access points (AP) are needed to avoid physical radio interference and optimize throughput. Mixed network scenarios where users have access to high-speed 5g from different carriers works the best. Future work will look into creating my own local network with off-the-shelf mesh wifi routers that come with fast roaming and multiple AP’s.4 I also found the initial centralized dome visualization app was kludgey at scale and needed attention. I initially developed the dome interface in SVG, considering that I could more easily position and animate waveforms. However, at 40 or more waveforms with SVG animation, the interface slows down immensely. I rewrote the visualization using an optimized Canvas 2d and no longer have issues. Additionally, there could be further work done to migrate the current browser-based interface to something like TouchDesigner that has more of a native and direct output to planetariums. Live interactive video is still a challenge for many planetarium infrastructure. Currently, I am able to set my dome interface within the browser at full screen and screen capture with Resolume (with circa .1 or .2 seconds of latency). It works well enough, but the added latency to the visuals can sometimes be noticeable. In some ways, the centralized modality of the system isn’t much of an improvement over a more straightforward feedback approach that would, for example, simply send an mp3 4It is important to me to be able to perform with this in no-bandwidth situations. Figure 7: Diagram of initial prototype. One Web Socket communicates timing. Another text Web Socket coordinates data between all clients. Each client sends and fetches audio (as whole files) via HTTP. or rtc stream to a server with delay and then loop that back into the audio with lowered volume. The advantage of using Web Audio and Web Timing Object is mainly to be able to visualize individual contributions and to more accurately time the samples in the distributed modality where the user’s mobile phones form the collective audio output. Even WebRTC has a latency of circa 30-150ms that gets amplified when considering the varied internal latencies of each mobile phone’s audio system. The Web Timing Object can compensate on the fly for the full latency of the system and keep devices within milliseconds of latency for output. There is still a slight choral effect, but very minimal. The kind of togetherness formed by this system is in some ways very trite - it’s just a bunch of people talking and making games out of spoken bits and pieces running acoustic ’laps’ around a track. However, the system is also potentially very volatile and complex. Our voices carry cultural and ethnic sub-harmonics. Our words convey our thoughts and beliefs. Where spaces exist for these human signals to mix, there is often a danger for explosive outcomes. There is no protection against hate speech, for example, in Audio Compost. It is this ’edge’ that gives the experience some weight and importance. To date, my observation of the system has mostly been on the level of play and inquiry, where users collectively explore rhythm and small verbal concepts. In one setting, users counted with numbers or collectively Figure 8: Diagram of second iteration. One Web Socket communicates timing. Another binary Web Socket coordinates data between all clients as well as sends and receives audio in circa 0.2s packets. listed the names of cheeses. In another setting in Wyoming (cowboy country), users listed western and cowboy items. More work could be done to coach and prompt the audienceparticipants in various ways. 5. FUTURE WORK Currently, the system lives as a usable experiment which can change drastically based on how an ’audience’ performs with it (or is instructed to perform with it). For the sake of usability and easy on-boarding of participants, the system is ’simple’ by design with only one push button feature to record. Its open ended nature is simultaneously inviting, but perhaps also unpunctuated. Often, I simply let the system run its course until the audience is ’done’ with it. This varies greatly on the context and can sometime end in underwhelming ways. I think future work could look into making it more performative with compositional structure, adjusting the GUI features for advanced users, practicing with it, and developing frameworks to guide the participants. To that end, I imagine live interactive radio and/or television could be an interesting use case for the system, potentially at a much larger scale. I imagine a system where one could input found sound in addition to voice. For example, if the system allowed one to search databases like freesound.org, archive.org, or the Wave Farm radio art archive and add canned content to a live running frippertronic system, thereby mixing archival, field, and found sounds to live voice and intonation, it could open the space of sound in more complex an nuanced ways. Additionally, I think it there could potential to combine live modes with other web-based live coding environments such as gibber, strudel, glicol, etc. I’m a firm believer that artistic software must be codesigned and co-evolved in collaboration with practitioners. In my opinion, the next step is to find performers to join me in developing experiments and techniques. I believe there is more to discover in voicing, call and response, and instigating audiences to participate with pre-planned compositional queues. The distributed and virtual nature of this system gives it special affordances that parallel and perturb the analogue counterpart. Technically speaking, the system might also find some speed and usability improvements with Chris Guttandin’s version of the timing server that works within WebRTC data exchange.[23] Figure 9: Browser-based dome UI using canvas 2D. In planetariums, we project map this image on the dome for direct visualization of the loop. 6. CONCLUSION Audio Compost is a collaborative frippertronic loop application that allows for free-form ludic play among multiple participants with voice. As a kind of speculative synthesis that changes based on audience mood and context, it lives in the web as an always ready 24/7 application that I use for impromptu social engagements as well as for more formal performance situations. It functions well as a small on-site interaction to build a sense of togetherness through poetic and interactive play around words, rhythm, and sonic interchange. With groups of 10-20 on a single local WiFi, the system works well. As the number of participants grow, more care must be placed in creating the network - especially in global locations where networking is cost-prohibitive. The next iteration will include a local WiFi with multiple AP’s so that it can perform more consistently. This paper describes the motivation for the project, prior art in participatory mobile music systems, and discusses the technical and aesthetic results. Future work will look to build timeline interfaces for integrating archival audio into the looped mix as well as scaling the system up even further for wide-area broadcast. 7. ACKNOWLEDGMENTS Thanks to Andrew Quitmeyer from Dinacon, Projekt Atol and Ljudmila at PIFcamp, Carol Sabbadini at Arte, Ciencia y Tecnologia de Idartes, the Bogot´a Planetarium, Alejandro Duque at CK:\WEB de Idartes, and Jane Crayton at the Wyoming Planetarium. Special thanks also go to MotionCorp at https://dev.mcorp.no/ for providing the stable timing server for this work. 8. REFERENCES [1] Arte, ciencia y tecnologia de idartes. https://www.idartes.gov.co/es/lineasestrategicas/arte-ciencia-tecnologia. [2] Ck:\WEB de idartes. https://ckweb.gov.co/. [3] Orf kunstradio. https://kunstradio.at. [4] Synapse, 1979. Summer Issue 1979. [5] Re-Inventing Radio. Revolver, Frankfurt am Main, Germany, July 2008. [6] Mediarecorder: start() method, 2025. https://developer.mozilla.org/enUS/docs/Web/API/MediaRecorder/start. [7] J. Allison and C. Dell. Aural: A mobile interactive system for geo-locative audio synthesis. In Proceedings of the International Conference on New Interfaces for Musical Expression, Ann Arbor, Michigan, 2012. University of Michigan. [8] M. Ananya, G. Essl, and M. Rohs. Microphone as sensor in mobile phone performance. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 185–188, Genoa, Italy, 2008. [9] J. J. Arango and D. M. Giraldo. The smartphone ensemble. exploring mobile computer mediation in collaborative musical performance. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 61–64, Brisbane, Australia, 2016. Queensland Conservatorium Griffith University. [10] I. Arntzen and F. Daoust. Timing object. https://webtiming.github.io/timingobject/. [11] I. M. Arntzen and N. T. Borch. Data-independent sequencing with the timing object: a JavaScript sequencer for single-device and multi-device web media. In Proceedings of the 7th International Conference on Multimedia Systems, MMSys ’16, pages 1–10. Association for Computing Machinery, 2016. https://dl.acm.org/doi/10.1145/2910017.2910614. [12] A. Barbosa. Displaced soundscapes: A survey of network systems for music and sonic art creation. 13:53–59. [13] A. Black. Hear-here: a choreographed peer-to-peer network for live participation on the radio. In Proceedings of the 12th International Audio Mostly Conference on Augmented and Participatory Sound and Music Experiences, pages 1–6. ACM, 2017. [14] T. Blaine and S. S. Fels. Contexts of collaborative musical experiences. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 129–134, Montreal, Canada, 2003. [15] T. Carson. Mesh garden: A creative-based musical game for participatory musical performance. In M. Queiroz and A. X. Sed´o, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 339–342, Porto Alegre, Brazil, June 2019. UFRGS. [16] E. Egozy and E. Y. Lee. *12*: Mobile phone-based audience participation in a chamber music performance. In T. M. Luke Dahl, Douglas Bowman, editor, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 7–12, Blacksburg, Virginia, USA, June 2018. Virginia Tech. [17] G. Elia and D. Overholt. Squidback: A decentralized generative experience, based on audio feedback from a web application distributed to the audience. In D. A. Mauro, S. Spagnol, and A. Valle, editors, SMC 2021: proceedings of the 18th Sound and Music Computing Conference: June 29th-July 1st 2021. Axea sas/SMC Network, 2021. [18] G. Essl. Speeddial : Rapid and on-the-fly mapping of mobile phone instruments. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 270–273, Pittsburgh, PA, United States, 2009. [19] C. Frisson, S. Dupont, J. Leroy, A. Moinet, T. Ravet, X. Siebert, and T. Dutoit. Loopjam: turning the dance floor into a collaborative instrumental map. In Proceedings of the International Conference on New Interfaces for Musical Expression, Ann Arbor, Michigan, 2012. University of Michigan. [20] A. F¨ orster, A. Uhde, M. Komesker, C. Komesker, and I. Schmidt. Loopboxes - evaluation of a collaborative accessible digital musical instrument. In M. Ortiz and A. Marquez-Borbon, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 30–39, Mexico City, Mexico, May 2023. [21] H. Grundmann. Art Telecommunication. 1984. [22] H. GRUNDMANN. Horizontal radio. Neue Zeitschrift f¨ ur Musik (1991-), 157(5):34–37, 1996. [23] C. Guttandin. The timing object - a pacemaker for the web. In J. Monschke, C. Guttandin, N. Schnell, T. Jenkinson, and J. Schaedler, editors, Proceedings of the International Web Audio Conference, WAC ’18, Berlin, Germany, September 2018. TU Berlin. [24] S. W. Lee and J. Freeman. echobo : Audience participation using the mobile music instrument. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 450–455, Daejeon, Republic of Korea, May 2013. Graduate School of Culture Technology, KAIST. [25] S. W. Lee, A. Srinivasamurthy, G. Tronel, W. Shen, and J. Freeman. Tok! : A collaborative acoustic instrument using mobile phones. In Proceedings of the International Conference on New Interfaces for Musical Expression, Ann Arbor, Michigan, 2012. University of Michigan. [26] G. Levin. Dialtones: A telesymphony. https://www. flong.com/storage/pdf/reports/dialtones report.pdf. [27] LJUDMILA. Pifcamp 2022 epifany, 2022. https://pif.camp/sl/2022 epifany. [28] J. Oh, J. Herrera, N. J. Bryan, L. Dahl, and G. Wang. Evolving the mobile phone orchestra. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 82–87, Sydney, Australia, 2010. [29] S. Park. Collaborative mobile instruments in a shared ar space: a case of arlooper. In R. Michon and F. Schroeder, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 190–195, Birmingham, UK, July 2020. Birmingham City University. [30] A. Quitmeyer. The digital naturalism conference, 2022. https://www.2022.dinacon.org/. [31] D. Ramsay and J. Paradiso. Grouploop: A collaborative, network-enabled audio feedback instrument. In E. Berdahl and J. Allison, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 251–254, Baton Rouge, Louisiana, USA, May 2015. Louisiana State University. [32] C. Roberts, A. Forbes, and T. H´ ’ollerer. Enabling multimodal mobile interfaces for musical performance. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 102–105, Daejeon, Republic of Korea, May 2013. Graduate School of Culture Technology, KAIST. [33] C. Roberts, G. Wakefield, and M. Wright. Mobile controls on-the-fly: An abstraction for distributed NIMEs. In Proceedings of the International Conference on New Interfaces for Musical Expression, Ann Arbor, Michigan, 2012. University of Michigan. [34] T. Roel Pacobahyba Pessanha, P. Martins, and J. Manzolli. A geographically distributed performance mediated by a co-creative agent over the web. 2024. [35] S. Salazar, A. Piepenbrink, and S. Reid. Developing a performance practice for mobile music technology. In T. M. Luke Dahl, Douglas Bowman, editor, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 59–64, Blacksburg, Virginia, USA, June 2018. Virginia Tech. [36] R. M. Schafer. The tuning of the world. A. A. Knopf, 1st ed edition, 1977. [37] H. Schellinx. a young person’s guide to frippertronics, 2002. http://www.harsmedia.com/Amphibious/Projects/sigsym.html. [38] G. Schiemer and M. Havryliv. Pocket gamelan: a pure data interface for mobile phones. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 156–159, Vancouver, BC, Canada, 2005. [39] N. Schnell, B. Matuszewski, J.-P. Lambert, S. Robaszkiewicz, O. Mubarak, D. Cunin, S. Bianchini, X. Boissarie, and G. Cieslik. Collective loops: Multimodal interactions through co-located mobile devices and synchronized audiovisual rendering based on web standards. In Proceedings of the Eleventh International Conference on Tangible, Embedded, and Embodied Interaction, pages 217–224. ACM, 2017. [40] T. Shaw, S. Piquemal, and J. Bowers. Fields: An exploration into the use of mobile devices as a medium for sound diffusion. In E. Berdahl and J. Allison, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 281–284, Baton Rouge, Louisiana, USA, May 2015. Louisiana State University. [41] V. Shepardson and T. Magnusson. The living looper: Rethinking the musical loop as a machine action-perception loop. In M. Ortiz and A. Marquez-Borbon, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 224–231, Mexico City, Mexico, May 2023. [42] A. Tanaka. Mobile music making. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 154–156, Hamamatsu, Japan, 2004. [43] B. Taylor. A history of the audience as a speaker array. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 481–486, Copenhagen, Denmark, 2017. Aalborg University Copenhagen. [44] B. Vann, J. Molina-Garcia, and A. Black. Pulse memorial: A web broadcasted multichannel sound installation exploring queer grief through audience participation. In SIGGRAPH Asia 2024 Art Papers, pages 1–6. ACM, 2024. [45] N. Weitzner, J. Freeman, S. Garrett, and Y.-L. Chen. massmobile -an audience participation framework. In Proceedings of the International Conference on New Interfaces for Musical Expression, Ann Arbor, Michigan, 2012. University of Michigan. [46] M. Wozniewski, N. Bouillot, Z. Settel, and J. R. Cooperstock. Large-scale mobile audio environments for collaborative musical interaction. In Proceedings of the International Conference on New Interfaces for Musical Expression, pages 13–18, Genoa, Italy, 2008. [47] A. Xamb´o and V. Goudarzi. The mobile audience as a digital musical persona in telematic performance. In Proceedings of the International Conference on New Interfaces for Musical Expression, The University of Auckland, New Zealand, jun 2022. [48] A. Xamb´o and G. Roma. Performing audiences: Composition strategies for network music using mobile phones. In R. Michon and F. Schroeder, editors, Proceedings of the International Conference on New Interfaces for Musical Expression, pages 55–60, Birmingham, UK, July 2020. Birmingham City University. [49] Zeitgleich (Project) and H. Grundmann. Zeitgleich: the Symposium, the Seminar, the Exhibition = das Symposium, das Seminar, die Ausstellung. Triton, 1994. OCLC: 38453413.