Creating Aesthetic Sonifications on the Web with SIREN
Tristan Peng, Hongchan Choi, Jonathan Berger
TL;DR
SIREN addresses the need for a general-purpose, accessible web platform for data sonification by leveraging parameter-mapping and a DAW-like interface to map datasets to a wide range of auditory parameters. The approach centers on a modular synthesizer schema implemented as JavaScript modules, enabling diverse synthesis methods (e.g., frequency modulation, formant, granular) with optional inline AudioWorklets. The paper details SIREN’s design, architecture, workflow, and concrete syntheses, highlighting its flexibility and extensibility through a clear module export convention and live data-driven control. It also discusses future directions, including integration with WebAssembly and community-sourced synthesizers to broaden the tool’s capabilities and reach.
Abstract
SIREN is a flexible, extensible, and customizable web-based general-purpose interface for auditory data display (sonification). Designed as a digital audio workstation for sonification, synthesizers written in JavaScript using the Web Audio API facilitate intuitive mapping of data to auditory parameters for a wide range of purposes. This paper explores the breadth of sound synthesis techniques supported by SIREN, and details the structure and definition of a SIREN synthesizer module. The paper proposes further development that will increase SIREN's utility.
