SoniChain

Real-time sonification of Binance market microstructure through an RNBO DSP engine.

SoniChain translates a live crypto order flow into a continuous, low-fatigue acoustic field. Five market metrics drive three noise-excited synthesis engines (subtractive, FM, physical-model waveguide) and a Schroeder–Moorer diffusion network, so that market state can be monitored peripherally — as a "sixth sense" — rather than read off a chart.

Watch the live demo on YouTube: https://youtu.be/KYNdKtD0t8s

Overview

SoniChain is a desktop-and-web application that converts a live Binance data stream into a real-time auditory display. The design objective is not merely to encode data as sound, but to do so under an explicit ergonomic constraint: the listener must be able to keep the stream running in the background of attention for prolonged sessions without listening fatigue. Every mapping is therefore chosen against a psychoacoustic rationale (spatial imbalance cues, interval-based pitch motion instead of glissando, perceptually motivated timbral modulation) rather than against a naive linear data-to-frequency rule.

The audio engine is authored in RNBO (Cycling '74) and exported to a WebAssembly/JS runtime; the host application is a SvelteKit 2 / Svelte 5 frontend that runs both as a web app and, via Tauri 2, as a native desktop shell. Market metrics are computed off the main thread in a TypeScript Web Worker and pushed to the DSP device through a strictly unidirectional control path.

Features

Five-metric market sonification — volatility, price, market_volume, density, maker_side, each mapped to a perceptually motivated DSP target.
Three switchable synthesis engines, theme-selected at runtime: an inharmonic subtractive glassarmonica, a dual-serial FM pad, and a Karplus–Strong / Jaffe–Smith bowed-string waveguide.
8-voice polyphony per engine, so price transients trigger discrete overlapping notes (harp-like decay overlap) instead of a continuous pitch sweep — no siren-effect glissando, reduced auditory fatigue.
Order-Flow-Imbalance spatialization — sell pressure to the left channel, buy pressure to the right, giving an immediate non-symbolic map of market lean.
Schroeder–Moorer stereo reverb with RT60 and damping driven by live market state.
Adaptive dynamic scaling — a self-narrowing peak-detection window keeps the mapping calibrated across regime changes without manual re-ranging.
Selectable engagement — a logarithmic market_volume sensitivity curve (Low / Mid / High) lets the user dial the display from a discreet macro-event alarm to full micro-structure immersion.
Built-in monitoring — an AnalyserNode-backed OutputScope for scope/meter feedback.
Cross-platform — pure web app in development, native desktop via Tauri 2.

Architecture & Data Flow

The signal path is strictly unidirectional: the worker computes metrics, the Svelte orchestrator forwards them as RNBO parameters, the DSP device synthesizes audio, and the Web Audio graph routes it to the output and to the monitoring scope. There is no feedback path from audio back into the data layer.

flowchart TD
    subgraph ACQ["Data Acquisition · Web Worker (TypeScript)"]
        BIN["Binance WebSocket<br/>aggTrade stream"]
        WORK["crypto.worker.ts<br/>microstructure metric computation"]
        BIN -->|"raw market events"| WORK
    end

    subgraph ORCH["Orchestration · Svelte 5"]
        PAGE["+page.svelte<br/>RNBO device host"]
    end

    subgraph DSP["RNBO DSP Device · @rnbo/js + DSP.export.json"]
        direction TB
        CTRL["Control Layer<br/>adaptive scaling · log volume curve<br/>maker_side to stereo pan (OFI)"]
        NOISE["Noise excitation<br/>amplitude from market_volume"]
        FREQ["scales + note_changer<br/>price delta to quantized interval"]
        subgraph ENG["Synthesis Engines · 8-voice poly · theme-selected"]
            direction LR
            GLASS["Glassarmonica<br/>dual inharmonic biquad BPF"]
            FM["FM Pad<br/>4-op dual-serial · ratio 1:1"]
            STR["Bowed String<br/>Karplus-Strong waveguide"]
        end
        REV["Schroeder-Moorer Reverb<br/>4 comb + 2 allpass per channel"]
        LIM["Brickwall Limiter"]
        CTRL --> NOISE
        CTRL --> FREQ
        NOISE --> ENG
        FREQ --> ENG
        ENG --> REV
        REV --> LIM
    end

    subgraph OUT["Audio Output + Monitoring · Web Audio API"]
        direction LR
        AC["AudioContext<br/>destination to speakers"]
        AN["AnalyserNode"]
        SCOPE["OutputScope.svelte<br/>scope / meters"]
        AC --> AN --> SCOPE
    end

    WORK -->|"postMessage TICK { 5 metrics }"| PAGE
    PAGE -->|"setRnboParam()"| CTRL
    LIM -->|"stereo out"| AC

Pipeline in words: Binance WebSocket → crypto.worker.ts → postMessage(TICK) → +page.svelte → setRnboParam() → RNBO device → Web Audio (AudioContext) → speakers + AnalyserNode → OutputScope.svelte.

Parameter Mapping

Each market metric modulates a specific DSP target. The mapping is the core of the design: it is where market semantics become perceptual attributes.

Market metric	DSP target	Perceptual intent
`maker_side` (Order Flow Imbalance)	Stereo pan — Bearish → L, Bullish → R	Spatial, pre-attentive map of market lean; keeps the stereo center uncluttered
`price` (discrete Δ via `note_changer`)	Pitch / interval — Δ > 0 → ascending, interval width ∝ \|Δ\|	Direction and magnitude of price movement, communicated as a musical step
`market_volume`	Noise excitation amplitude (log curve; Low / Mid / High)	Event intensity; user-selectable immersion depth
`density`	Filter cutoff / brightness, comb gain, detune rate	"Proximity" and clarity; tighter trade flow reads as a fuller, closer, brighter sound
`volatility`	Decay / feedback / loop gain, reverb RT60	Acoustic "memory": turbulent markets leave a persistent, reverberant tail

The RNBO Patch (DSP)

The DSP is a single RNBO export (DSP.export.json) composed of a control/routing layer, three polyphonic synthesis engines, and a diffusion network. The four signal-generating/processing cores are implemented as RNBOScript codeboxes; everything below is documented from those sources, which are the source of truth for constants and topology.

A recurring engineering theme across all codeboxes: per-sample one-pole smoothing (τ ≈ 5–10 ms) on every control parameter to suppress zipper noise on the discretized, WebSocket-rate updates, plus explicit safety clamps and denormal flushing.

Control & Routing Layer

Ingest & master output. The layer receives the five metrics, applies the spatialization and scaling logic, and feeds a brickwall limiter upstream of the stereo output.
Adaptive scaling subpatch. Raw values are normalized to [0.0, 1.0] by a peak-detector with an adaptive threshold whose temporal window progressively narrows to track macroscopic range shifts — so a regime change (e.g. sideways → crash) re-ranges the display automatically, without manual recalibration.
market_volume sensitivity is shaped by a logarithmic transfer curve with user-selectable Low / Mid / High settings, trading off macro-event salience against micro-structure detail.
Spatialization. maker_side drives the stereo panorama as a direct read-out of Order Flow Imbalance: sell pressure (Bearish) → left, buy pressure (Bullish) → right.
Frequency assignment (scales + note_changer). scales emits pre-quantized frequency arrays, constraining tonal output to the selected scale. note_changer computes the discrete derivative of the last two price ticks: a positive delta triggers ascending harmonic motion, with interval width proportional to |Δ|. This module also drives the engine switch based on the active UI theme.

Engine A — Glassarmonica (Subtractive Resonator)

High-Q subtractive synthesis emulating the inharmonic spectrum of a glass harmonica.

Topology — two parallel resonant biquads (Direct Form I), implemented as RBJ constant-0 dB-peak band-pass sections (α = sin ω / 2Q, b0 = α/a0, b2 = −α/a0, a1 = −2cos ω/a0, a2 = (1−α)/a0). The second filter is tuned to f0 × 2.756, an inharmonic partial ratio that gives the glass-like timbre. Coefficients are recomputed at sample rate to follow modulation.
Self-sustaining resonance — a cross-channel feedback loop (input_L = in1 + fb_R·feedback, input_R = in2 + fb_L·feedback) sustains the ring across the stereo pair.
Excitation — pre-weighted white noise, low-passed as a function of density.
Modulation — density modulates the filter cutoff; volatility controls the feedback-loop gain. The codebox exposes f0, Q, and feedback as inlets; the metric-to-inlet mapping is applied upstream.

Engine B — FM Pad (Dual-Serial Frequency Modulation)

A timbrally stable, noise-excited "warm bell" pad.

Topology — four operators in two independent serial chains, Op2 → Op1 (L) and Op4 → Op3 (R), at a carrier:modulator ratio of 1:1, yielding a fully harmonic spectrum (Chowning 1973: an integer ratio places every sideband on an integer multiple of f0).
Excitation model — excitation is 100% external: in1/in2 carry market_volume-weighted noise. The noise is never summed into the output; instead it drives the FM modulation-index envelope and the feedback path, exactly as the excitation drives the delay line in a Karplus–Strong string. When the noise falls to zero, energy already in transit in the feedback path continues to decay by the coefficient from in4 — a physical-tail behavior, not a hard cutoff.
Modulation index — a stable floor (0.35) plus an envelope-driven component, hard-clamped to a ceiling of 2.55 (0.35 + 2.2). The range is deliberately conservative: at a 1:1 ratio, indices beyond ~4–5 collapse the spectrum into a dense buzz.
Decoupled envelopes (the key fix) — two one-pole followers share the same noise source but serve perceptually distinct tasks. A 15 ms follower drives amplitude; a slower 40 ms follower drives the modulation index. The slower stage sits below the lower roughness boundary (Zwicker & Fastl: flutter to ~20 Hz, roughness 20–300 Hz), removing the residual band that a single 15 ms follower would otherwise pass into the index path as PM-index jitter — i.e. audible grain / spurious sidebands from phase modulation of the carrier at a perceptible rate. (The artifact was most audible in the reverb return, because the output leaky integrator and the reverb combs — both leaky-integrator topologies — accumulate and prolong it.)
Stereo width — symmetric detune of up to ±1.5 Hz between the two carrier chains (≈ 3 Hz beat) produces slow chorus-like "breathing," not perceived inharmonicity.
Decay — a leaky integrator on the output models a Karplus–Strong-like physical tail post-excitation.
Modulation — density governs the noise low-pass and the detune rate; volatility controls the leaky-integrator feedback.
Disclosure (from source): no oversampling in this codebox; the conservative index keeps FM products clear of Nyquist. If f0 is pushed beyond ~1–2 kHz in sound-design, consider 2× oversampling at the patcher level.

Engine C — Bowed String (Resonant Waveguide)

A continuously excited physical model (Karplus–Strong / Jaffe & Smith 1983) — the only string-bodied voice in the palette.

Topology — a tuned delay line (length sr / f0, fractional, with 2-tap linear FIR interpolation) and filtered feedback. The external noise is injected continuously into the loop (project constraint: the excitor remains the external noise).
Root-cause stability fix (Revision 5). The original one-pole loop filter had unity DC gain for any feedback coefficient, so that coefficient controlled only color, not decay — and in DC the loop degenerated into a perfect integrator over the delay period, accumulating any DC component of the excitation into unbounded growth and a fixed-frequency buzz independent of f0. The canonical fix introduces three explicit stages:
1. an explicit loop gain < 1 (DECAY_MAX = 0.995), separated from damping — this is the true decay, controlled by volatility;
2. a first-order DC blocker inside the loop (y[n] = x[n] − x[n−1] + R·y[n−1], R = 0.999 ≈ 7–8 Hz cutoff @ 48 kHz), nulling the loop's DC gain so no DC ever circulates in the delay;
3. energy normalization of the injection (× (1 − loop_gain)): at high Q the comb's resonant peak ≈ 1/(1 − loop_gain) ≈ 200, so without compensation a strong injection (high market_volume) would saturate and clip. With it, steady-state amplitude tracks the injection level independently of loop_gain, while loop_gain controls only tail duration.
Analytic guarantee — loop transfer = HP(z)·loop_gain·LP(z) with |HP| ≤ 1, |LP| ≤ 1, loop_gain < 1 ⇒ loop gain < 1 at all frequencies, and → 0 in DC. Unconditionally stable.
Damping — a fixed-coefficient averaging filter 0.5·(x[n] + x[n−1]) (zero at Nyquist), classic K–S, frequency-invariant. The two excitation stages — brightness LP and resonant comb, both density-controlled — are strictly feedforward, outside the loop.
Transitions — a linear crossfade (XFADE_SAMPLES = 480) on real fundamental jumps eliminates phase clicks from delay-buffer discontinuity.
Disclosure (from source): not empirically verified in this context (no RNBO runtime available at authoring time). Linear delay interpolation has a less flat high-frequency phase than an allpass (Välimäki & Laakso 2000) but is intrinsically stable.

Diffusion — Schroeder–Moorer Reverb

A dedicated stereo spatial processor for the FM and waveguide voices.

Topology — 4 parallel comb filters + 2 cascaded allpass per channel (Schroeder 1962; Moorer 1979), with in-loop damping. RT60→gain derivation follows Smith, Physical Audio Signal Processing (CCRMA) — the same reference used by the waveguide module.
Decorrelation — comb lengths are coprime and decorrelated L/R (Dattorro 1997) to maximize echo density without audible periodicity. Lengths are specified in samples at 48 kHz and auto-rescaled by sr / 48000 to preserve the temporal ratios.
Decay law — Schroeder's g = 10^(−3·d / (RT60·sr)) gives exactly −60 dB at t = RT60. A one-pole LPF in each comb loop (damping) models the faster HF absorption of real absorptive materials (Kuttruff, Room Acoustics).
Allpass diffusion — y[n] = −g·x[n] + x[n−d] + g·y[n−d], g = 0.6 (within [0.5, 0.7] for guaranteed stability and density without metallic coloration).
Wet/dry — a deliberately linear mix. Because the reverb is additive and the dry/wet sum is decorrelated, a linear crossfade does not produce the loudness dip of an equal-power crossfade between correlated signals — so linear is the correct choice here, not an omission.
Modulation — volatility scales RT60; density controls the damping (brightness).

Sound Design Rationale (Stylistic Choices)

The art of listening to data: psychoacoustic and UX choices in market sonification.

Translating a financial data stream into sound is a UX problem, not just a technical one. The primary objective is to let the brain process market complexity in the background — intuitively, and without the listening fatigue typical of long sessions. Every market parameter is therefore bound to a deliberate psychoacoustic metaphor.

Logical spatialization — the weight of the market. maker_side, the order-flow imbalance, governs the stereo panorama. Sell pressure (Bearish) is steered left, buy pressure (Bullish) right. The listener does not decode a complex timbre; they physically feel which way the market is leaning. The hard L/R separation also keeps the stereo center uncluttered, in line with the visual interface.

Price — intervals, not glissando. The most critical datum uses a universal auditory metaphor: a rise in price raises pitch, a fall lowers it, and larger price moves produce wider harmonic leaps, communicating intensity instantly. Crucially for the listener's ear, the system avoids glissando — the continuous "siren" slide. By assigning 8 polyphonic voices per engine, price transients trigger discrete notes that overlap fluidly, producing an organic, harp- or natural-reverb-like field. This removes the sense of artificiality and makes prolonged listening restful.

Timbre & texture — hearing density and volatility. density controls brightness and filter cutoff: a denser market sounds closer, clearer, more present; in the FM pad it also introduces slight detuning whose slow beats read as "thickness" and "breath" rather than mistuning. volatility acts on decay and reverb tail: psychoacoustically it models memory / turbulence — in a highly volatile market, events leave a persistent acoustic wake and fuse into a reverberant, chaotic ambience; as volatility drops, the acoustic space dries out, becoming intimate, precise, defined.

Adaptability & cognitive load. A display that screams at every spike quickly becomes intolerable. Adaptive logic continuously re-ranges the value excursion in the background, so a regime change re-scales smoothly rather than clipping. The logarithmic market_volume sensitivity curve then lets the user choose their level of involvement: Low acts as a discreet alarm for macro-events only; High allows total immersion in the market's micro-pulse.

In short, the sound design moves past literal data playback toward sensory ergonomics — intensity becomes harmonic amplitude, imbalance becomes physical space, turbulence becomes acoustic persistence — yielding a monitoring instrument that does not fatigue, but integrates as a genuine sixth sense for market behavior.

Tech Stack

Layer	Technology
Frontend / orchestration	SvelteKit 2, Svelte 5 (`+page.svelte` as orchestrator)
Audio engine	RNBO Web export via `@rnbo/js` 1.3.4, loaded from `static/DSP.export.json`
Data layer	TypeScript Web Worker (`crypto.worker.ts`) — Binance WebSocket + metric computation
Audio I/O & monitoring	Web Audio API directly — `AudioContext` + `AnalyserNode` (no wrapper)
Build tool	Vite 6, with a custom `sync-dsp` plugin
Desktop shell	Tauri 2 (Rust); the frontend also runs as a pure web app in development
Dev server	Fixed port 1420 (`strictPort: true`)

Project Structure

SoniChain/
├── src/
│   ├── routes/
│   │   ├── +page.svelte          # Orchestrator: hosts the RNBO device, wires worker → setRnboParam
│   │   └── +layout.ts            # Disables SSR globally (SPA mode for Tauri)
│   ├── components/
│   │   ├── AudioControls.svelte  # Volume fader, asset / scale / sensitivity selectors, play/pause
│   │   ├── CryptoChart.svelte    # Price history chart with musical-note overlay
│   │   └── OutputScope.svelte    # AnalyserNode-backed scope / goniometer / dB meters
│   ├── RNBO/
│   │   ├── DSP1.json             # Authoring-side RNBO export (source of truth for the DSP)
│   │   ├── DSP1.license
│   │   └── dependencies.json
│   ├── crypto.worker.ts          # Binance WebSocket + microstructure metric computation
│   ├── themes.ts                 # Three themes (graphite / slate / bone) as CSS vars + canvas hex
│   ├── help.ts                   # Contextual help-bubble store + Svelte `help` action
│   └── app.html                  # SPA HTML shell
├── static/
│   └── DSP.export.json           # Runtime device, synced from src/RNBO/DSP1.json — do not edit
├── src-tauri/
│   ├── src/
│   │   ├── main.rs               # Tauri entry point
│   │   └── lib.rs                # Tauri commands
│   ├── tauri.conf.json           # Window config: 800×600, title "SoniChain"
│   ├── Cargo.toml
│   ├── build.rs
│   └── capabilities/             # Tauri v2 security capabilities
├── vite.config.js                # Vite 6 config + sync-dsp plugin + strictPort 1420
└── package.json

Getting Started

Prerequisites

Node.js ≥ 20 (required by Vite 6 / SvelteKit 2) and a package manager (npm / pnpm).
Rust (stable toolchain) and the platform-specific Tauri 2 prerequisites — see the Tauri prerequisites guide — only needed for the desktop build.
A network connection for the Binance WebSocket feed.

Installation

git clone https://github.com/BitResonant/SoniChain.git
cd SoniChain
npm install

Development

Run the frontend as a pure web app (dev server on the fixed port 1420):

npm run dev

Run inside the Tauri desktop shell:

npm run tauri dev

Build

Web build:

npm run build
npm run preview      # serve the production build locally

Native desktop bundle:

npm run tauri build

Type-checking

npm run check        # svelte-kit sync + svelte-check (strict TypeScript)
npm run check:watch  # same, in watch mode

This is the only automated correctness gate — run it after any TypeScript or Svelte change.

The DSP Sync Workflow

The RNBO patch is authored as src/RNBO/DSP1.json but consumed at runtime as static/DSP.export.json. A custom Vite plugin, sync-dsp, keeps them in lockstep:

On buildStart — copies src/RNBO/DSP1.json → static/DSP.export.json.
During dev — watches src/RNBO/DSP1.json; on change, re-copies and triggers a full page reload so the new device loads immediately without a manual restart.

Workflow for editing the DSP: edit the patch in RNBO, re-export to src/RNBO/DSP1.json — the dev server picks up the change automatically (or run npm run build for a production rebuild). Do not edit static/DSP.export.json by hand; it is a generated artifact.

Data Source

Market data is sourced from Binance over a WebSocket connection, consumed in crypto.worker.ts. The worker computes the five microstructure metrics (volatility, price, market_volume, density, maker_side) off the main thread and pushes them to the orchestrator via postMessage TICK events. No credentials are required for public market streams.

References

The DSP design is grounded in the following literature, as cited throughout the codeboxes:

Chowning, J. M. (1973). The Synthesis of Complex Audio Spectra by Means of Frequency Modulation. Journal of the Audio Engineering Society, 21(7).
Karplus, K., & Strong, A. (1983). Digital Synthesis of Plucked-String and Drum Timbres. Computer Music Journal, 7(2).
Jaffe, D. A., & Smith, J. O. (1983). Extensions of the Karplus–Strong Plucked-String Algorithm. Computer Music Journal, 7(2).
Schroeder, M. R. (1962). Natural Sounding Artificial Reverberation. Journal of the Audio Engineering Society, 10(3).
Moorer, J. A. (1979). About This Reverberation Business. Computer Music Journal, 3(2).
Dattorro, J. (1997). Effect Design, Part 1: Reverberator and Other Filters. Journal of the Audio Engineering Society, 45(9).
Välimäki, V., & Laakso, T. I. (2000). Principles of Fractional Delay Filters. IEEE ICASSP.
Zwicker, E., & Fastl, H. Psychoacoustics: Facts and Models. Springer.
Kuttruff, H. Room Acoustics. CRC Press.
Smith, J. O. Physical Audio Signal Processing. Online book, CCRMA, Stanford University.

License

This project is distributed under the terms of the LICENSE file in the repository root.

SoniChain — BitResonant

Name		Name	Last commit message	Last commit date
Latest commit History 36 Commits
.github/workflows		.github/workflows
.vscode		.vscode
src-tauri		src-tauri
src		src
static		static
.gitignore		.gitignore
CLAUDE.md		CLAUDE.md
README.md		README.md
icon_source.png		icon_source.png
package-lock.json		package-lock.json
package.json		package.json
svelte.config.js		svelte.config.js
tsconfig.json		tsconfig.json
vite.config.js		vite.config.js

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

SoniChain

Table of Contents

Overview

Features

Architecture & Data Flow

Parameter Mapping

The RNBO Patch (DSP)

Control & Routing Layer

Engine A — Glassarmonica (Subtractive Resonator)

Engine B — FM Pad (Dual-Serial Frequency Modulation)

Engine C — Bowed String (Resonant Waveguide)

Diffusion — Schroeder–Moorer Reverb

Sound Design Rationale (Stylistic Choices)

Tech Stack

Project Structure

Getting Started

Prerequisites

Installation

Development

Build

Type-checking

The DSP Sync Workflow

Data Source

References

License

About

Uh oh!

Releases 1

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

SoniChain

Table of Contents

Overview

Features

Architecture & Data Flow

Parameter Mapping

The RNBO Patch (DSP)

Control & Routing Layer

Engine A — Glassarmonica (Subtractive Resonator)

Engine B — FM Pad (Dual-Serial Frequency Modulation)

Engine C — Bowed String (Resonant Waveguide)

Diffusion — Schroeder–Moorer Reverb

Sound Design Rationale (Stylistic Choices)

Tech Stack

Project Structure

Getting Started

Prerequisites

Installation

Development

Build

Type-checking

The DSP Sync Workflow

Data Source

References

License

About

Topics

Resources

Uh oh!

Stars

Watchers

Forks

Releases 1

Contributors

Uh oh!

Languages