Session 2 — Paper Presentations

Ambisonics: Introducing an Accessible, Distributed Approach

Thomas Rushton, Romain Michon, Tanguy Risset, Clemens Wegener and Pierre Lecomte

Keywords: Ambisonics, Accessibility, Distributed DSP, Microcontroller, Networked audio

Abstract

This paper presents an accessible, distributed approach to higher order ambisonics using a network of low-cost microcontroller-based devices. Leveraging recent advances in embedded computing, networked audio, and precise time synchronization, the system offers an alternative to conventional centralized architectures that typically require expensive multichannel interfaces and high-performance digital signal processing hardware. The proposed architecture combines an audio plugin for soundfield design with a modular network of ambisonics processing units, which, exploiting the parallelisable nature of ambisonics decoding, are each responsible for a subset of loudspeaker outputs. Synchronization is achieved through the IEEE 1588 Precision Time Protocol, complemented by adaptive clock correction to maintain sub-microsecond alignment between devices. A third-order ambisonics implementation demonstrates the viability of the approach, with potential scalability to higher orders and larger arrays. Overall, this work contributes to the development of frugal, distributed solutions for immersive audio, expanding access to ambisonics-based systems for a wide range of applications.

 

Ambiviz: A Python toolbox for Ambisonic audio and 360-degree video visualization

Jinyue Guo, Maham Riaz and Alexander Refsum Jensenius

Keywords: spatial audio, audio visualization, Python toolbox

Abstract

This paper introduces ambiviz, A Python toolbox for Ambisonics and 360-degree video visualization. While many audio plug-ins and proprietary softwares support many forms of spatial audio visualization, we found two main issues when using them for research purposes: (1) existing plug-ins and softwares do not support a programming interface, making it difficult to either batch process datasets or modify the visualizations according to different needs; (2) current visualization graphs cannot represent stationary indoor environments well. To fill this gap, we propose Ambiviz, an open-source Python toolbox for batch processing, easily modifiable via the matplotlib backend, and including custom visualizations for indoor environments. Besides common visualizations such as the audio energy map, Ambiviz also supports the anglegram and directogram to show different spatial-temporal aspects in the Ambisonics recording. The source code for ambiviz is available at https://github.com/fisheggg/ambiviz.

 

 

From SPRAWL to MESH: Networked Music Systems for Touring and Stage Integration

Henry Windish and Henrik von Coler

Keywords: Networked Music Performance, Distributed Music Systems, Electroacoustic Ensemble, Portable Performance Systems

Abstract

MESH is a wireless, networked instrument for music performances and installations. The next development of the SPRAWL system, MESH consists of a set of interaction nodes that are wirelessly dispersed throughout a space. The system distributes sound generation, control, and interaction across compact, self-contained nodes, each capable of operating independently while remaining networked through OSC-based communication. Designed for use as a performance instrument and for interactive installations, the MESH system is portable and can be set up quickly. The design of MESH is motivated by prior experience with SPRAWL V2, a wired distributed ensemble system, and reflects a shift toward lightweight, decentralized infrastructure. This paper outlines the system’s hardware and software architecture and documents its use across domestic and international performances, installations, and workshops. demonstrating the systems portability. The system has traveled domestically and abroad for installation, performance, and workshop opportunities, demonstrating the system's portability. This evaluation contributes a practical and scalable approach to networked music performance, emphasizing portability, reduced technical overhead, and adaptability to diverse venues.

 

Audience as Performer: A Smartphone Web-Application For Real-Time Sound Contribution

Davor Vincze and Maurice Oeser

Keywords: Technology-Mediated Audience Participation, Smartphone as Musical Instrument, Participatory Performance, Audience Co-Creation, Real-Time Sound Recording

Abstract

This paper presents a participatory performance system that enables audience members to contribute directly to the sonic material of a live electronic music performance using their smartphones. The system builds on earlier iterations developed for projects Freedom Collective and Exo Signals, where smartphones functioned as a distributed speaker array and interactive control interface. The current iteration introduces a new module, allowing participants to record short sound bytes that are integrated into the performance in near real time. This development represents a shift from audience interaction through control to audience participation through sound generation. The system was evaluated in three performances at the Australian National University (Canberra), Emory University (Atlanta), and CCRMA (Stanford), followed by a post-concert survey examining usability, enjoyment, and perceived agency. The findings suggest that enabling audience members to contribute their own recorded material increases perceived engagement and sense of participation in the performance.

 

Exploring Methods for Measuring Audience Body Motion in Concert Halls

Hugh Alexander von Arnim, Finn Upham and Alexander Refsum Jensenius

Keywords: Motion Analysis, Live Concert Research, Thermography, IR, Accelerometery, Quantity of Motion, Audience Motion

Abstract

Measurement of audience members’ body motion in ecological concert settings must navigate environmental constraints, access to audience members, privacy considerations, and behavioural norms of the performance event. This study compares five distinct technologies for capturing motion from “audience member” participants under simulated symphonic concert conditions: visible-range imaging (RGB), near-infrared imaging (IR), thermal imaging, thermal data, and torso-mounted accelerometry. Seven quantity of motion (QoM) time series were generated from these measurements, each cleaned and normalised to reveal similarities and differences in the information they provide regarding four concert-typical behaviours: Large Actions such as posture changes, Applause, Stillness, and Respiration. Results show comparable, though not identical, quantity of motion measurements across devices in this simulated concert setting. We discuss the advantages and disadvantages of the various signals, share recommendations on key measurement parameters, and discuss scaling these motion capture methods for larger audiences in real concert halls.

 

Organic Reflections: A Stochastic Framework For Real-Time Audiovisual Composition Across Ar, Vr, And Projection

Deniz Caglarcan

Keywords: audiovisual composition, Visual Music, parameter mapping, Unreal Engine 5, Algorithmic Composition

Abstract

This paper presents a real-time audiovisual composition framework in which sound and image are shaped by the same stochastic decisions, producing structurally coupled output without post-hoc synchronisation. Built entirely within Unreal Engine 5, the system uses a weighted-probability scheduler to simultaneously select from five heterogeneous sound engines, implemented in MetaSounds using analogue-modelled synthesis, granular processing, and sample playback, and route the resulting amplitude envelopes to Niagara GPU particle systems, shaping mesh speed, birth rate, force, and scale in real time. The framework externalises compositional decision-making into a generative system: the composer authors a possibility space of gestural and textural audiovisual behaviours, and the scheduler navigates it in real time, producing structurally guaranteed synchresis through shared kineto-morphological evolution. The system deploys from a single project to three presentation modes, mobile augmented reality, virtual reality on Meta Quest, and projection, preserving the parametric audiovisual bond while reshaping the audience's spatial relationship to the work. The framework is demonstrated through the generative audiovisual work Organic Reflections.

 

 

Xplosive Collision: Designing a Noise-Oriented Networked Music System Based on Dyadic, Interdependent Game Mechanics

Luca Morino, Francesco Ardan Dal Rì, Fabio Cifariello Ciardi and Nicola Conci

Keywords: networked music system, mobile music system, games, participatory music-making, noise music, experimental music

Abstract

This paper presents Xplosive Collision (XC), an interactive networked music system for two players. Implemented as a web application for smartphone browsers over a wireless local network, the system interconnects participants’ devices and uses competitive game mechanics to structure dyadic interaction within a noise/experimental music paradigm. Game objectives have been grounded on timbral analysis of the generated sound. The system has been evaluated through an exploratory study with seven pairs of players, primarily based on questionnaire data collected through a post-session questionnaire, supported by system logs and optional open comments. The study examines how participants interact with the system and with one another, and how they engage with its game mechanics and sonic outcomes. Results indicate that participants effectively relied on sound to orient themselves within the system and game mechanics to guide their actions, and remained overall focused on the game objective.

 

 

From Spatialization to Spatial Structure: A Spatial Grammar for Electroacoustic Composition

Marcela Pavia

Keywords: Spatial Audio, Electroacoustic Composition, Spatial Grammar, Sound Spatialization

Abstract

Spatialization has become a central dimension of electroacoustic music, yet compositional approaches often remain closely tied to specific spatial rendering technologies such as Wave Field Synthesis, Ambisonics, or multichannel diffusion. This dependency raises a fundamental compositional challenge: how can spatial structures be conceived in a way that remains meaningful across different spatial audio systems? This paper proposes the concept of Spatial Grammar as a compositional framework that separates spatial structure from rendering technologies. Drawing on spectromorphological theory, models of spatial perception, and concepts of perceptual articulation, the framework organizes spatial processes across multiple levels—parametric, functional, syntactic, and gestural—through which spatial relationships contribute to musical form and perceptual organization. The approach is developed through the electroacoustic work Risonanze, originally composed in a Wave Field Synthesis environment, where spatial parameters directly influence processes of sound generation and transformation. The translation of this work into an eight-channel configuration highlights the limitations of system-dependent spatial composition. To address this issue, the paper presents a compositional tool that enables the translation of spatial trajectories across rendering environments through intermediate data representations. More broadly, the study proposes a compositional paradigm in which spatial structure is decoupled from spatial rendering, allowing space to function as an active and transferable dimension of musical organization.

 

Dynamic Game Audio Analysis in The Context of Horror Games

Victor Schmaedecke, Genaro Costa, Claudio Carvilhe and Carlos Nascimento Silla Junior

Keywords: Game Audio, Music Analysis, Dynamic Audio

Abstract

This article aims to highlight the relationship between ludonarrative contexts in games and their use of Adaptive Music. To better suit the interactive nature of games, Adaptive Music functions as a way for the soundtrack to respond to the game environment. The research presents the development and application of a new structured analysis model, designed for observing Adaptive Music across a large number of games, enabling a descriptive analysis of the collected data. As part of the model, the study provides a categorical definition of the ludonarrative elements that compose scenes in these games and identifies 30 contexts in which Adaptive Music was used. The results indicate that 76\% of the 221 unique Adaptive Music applications in the 10 games studied occur in only 6 main identified contexts. The initial search for a relationship between implementation techniques and contexts suggests that each context shows different tendencies but tends to favor up to two of the three Adaptive Music techniques — Horizontal Re-sequencing, Vertical Layering, and Run-Time Mixing — with a greater number favoring Horizontal Re-Sequencing. Additionally, it was observed that only 29 applications used multiple techniques simultaneously and that most applications are never reused throughout the remainder of the game.

 

 

A Real-Time Procedural Music Generation System Based on Affective Content of Low-Level Video Features

Filippo Giabelli, Federico Avanzini and Giorgio Presti

Keywords: Adaptive Soundtrack, Video to Music, Affective Computing

Abstract

This paper presents a proof of concept for a software architecture designed to generate real-time procedural music based on the affective content of arbitrary video streams, utilizing low-level video features. We achieved this objective by implementing a demonstrative application developed entirely within the Unreal Engine 5.4 environment. The system utilizes MetaSounds for music generation and the Pleasure-Arousal-Dominance (PAD) model to map affective values from imagery to music. Furthermore, it employs asynchronous computation to optimize processing load, ensuring suitability for real-time applications. The work has been validated through performance measurements and empirical testing.