Paper Session 2b: AI & Music

Session Chair: Paulo Chagas

Paper abstracts

Hiroshi Yamato, OrbitScore: “A Domain-Specific Language for Polymetric Live Coding Based on Multilayered Temporal Structures”

This paper presents OrbitScore, a domain-specific language (DSL) for live coding polymetric rhythm patterns based on the theory of Multilayered Temporal Structures (MLTS). While existing live coding languages such as TidalCycles and Sonic Pi provide rich pattern manipulation capabilities including polyrhythmic support, OrbitScore offers an intuitive syntax where the beat(n by m) notation directly
maps to the theoretical 4:(n/4) framework, enabling each sequence to maintain its own meter and allowing performers to create intricate polyrhythmic textures in real-time. The system integrates with SuperCollider for low-latency audio synthesis and provides a declarative, method-chaining syntax designed for live performance. We describe the theoretical foundation, DSL design, implementation architecture, and demonstrate the system’s capabilities through a live coding performance. Our contribution lies in bridging the gap between the theoretical framework of Multilayered Temporal Structures and practical live coding tools, making polymetric expressions accessible to performers.

Piero Poli: “Dancing Cabiria: An hyper-environment study through corpus-based techniques”

This paper introduces the concept of hyper-environment — an additional spatial layer superimposed on the choreographic space, where physical movement becomes a means
of navigating and activating pre-analyzed sound materials. The work examines Dancing Cabiria, a reenactment in four scenes from Giovanni Pastrone’s silent film Cabiria (1914), as a case study to explore performative hyper-environments that employ corpus-based synthesis techniques within a virtual reality framework. Through the use of motion-tracking suits, four choreographies are performed, each one by four dancers whose movements are translated into sound via audio corpora distributed throughout the virtual space surrounding each performer. Each choreography outlines different uses and configurations of this hyper-environment, and allow for the discussion of compositional and instrumental issues such as the scale and density of the corpora, the relationship that emerges between movements width, corpus dimensions, and virtual space volume, and the role of real-time feedback in the design of hybrid instruments for performers.

Minami Kojima, Takayuki Itoh and Rafael Ramirez: “Factor Analysis of Similarity in the Same Orchestral Piece”

There have been conventional research comparing multiple performances of the same classical music piece, where many of them primarily discussed the similarity between performances but rarely delved into why those performances are similar. Furthermore, conventional similarity analysis has dealt with the similarity of single features in classical music, such as timbre, tempo, and loudness. However, since actual audiences perceive these features compounded without separating them, relying solely on single features is insufficient to fully represent musical style. To overcome this challenge, our study uses a deep learning model (VGGish, specialized in audio feature extraction) that captures high-level timbral and textural features, in addition to acoustic features (timbre, loudness, tempo), targeting the same orchestral piece. Subsequently, based on those features, we define five grouping criteria: (1) same orchestra, (2) same conductor, (3) same country of orchestra, (4) same nationality of conductor, and (5) same teacher of the conductor. We then evaluate the clustering performance for each criterion. A group exhibiting high clustering performance suggests mutual similarity among the performances within that group, leading us to conclude that the corresponding criterion represents a major factor influencing performance style. The results in this paper show that conductor identity consistently yields
the strongest clustering for tempo-related features, while orchestra identity dominates timbral similarity in specific movements. We demonstrate that metadata-driven factors explain similarity beyond purely perceptual or affective similarity measures.