A resilient web framework for orchestrating massive distributed audio performance over audience smartphones — built for cellular networks, validated with a Guqin and 421 phones.
When hundreds of smartphones must function as a synchronized speaker array in one venue, the congestion and jitter of high-density 4G/5G networks demand an architecture that prioritizes connection survival over bidirectional interactivity.
Qianji adopts a resilience-first design: a unidirectional Server-Sent Events broadcast for command, decoupled from a stateless HTTP clock-synchronization channel. A multi-stage filtering pipeline (outlier rejection → RTT-ranked sample selection → EMA smoothing) keeps the collective instrument coherent even when individual connections wobble. We validate the framework through The Discourse of an Instrument, and a Thousand Machines, a piece that extends a traditional Guqin with an audience array of up to 421 smartphones.
The Discourse of an Instrument, and a Thousand Machines treats the audience array not as independent instruments but as the Guqin's Resonant Body — a distributed qinzhuo that receives and extends the zither's voice across the hall. Spatial gestures (sweeping waves, granular clouds) are pre-composed offline as grayscale video and rendered into per-seat amplitude envelopes via the Video-to-Volume workflow.
xianyuqianji — Hangzhou, live recording
@inproceedings{qianji2026,
title = {Qianji: A Resilient Framework for Orchestrating
"A Thousand Machines" in Distributed Performance},
author = {TBD},
booktitle = {Proceedings of the International Conference on
New Interfaces for Musical Expression (NIME)},
year = {2026},
address = {London, United Kingdom}
}