Process — 02 of 04

02

Aggregation
the individual
dissolves into the field

Thirty streams of physiological data converge into a single collective state — a portrait of the group rather than a record of its parts.

Overview

Group-level dynamics take precedence over individual patterns.

Once signals are sensed, they must be fused. Aggregation is where individual physiological streams are synthesized into a coherent representation of the group's shared state — a single data structure that captures the collective without erasing the nuance of its constituent parts.

The system does not average. It weighs, clusters, and measures coherence. The result is a three-dimensional emotional field described by valence, arousal, and coherence — the core coordinates of collective experience.

3D
Emotional space — valence, arousal, and coherence as primary coordinates
<100ms
Aggregated state made available to the translation layer
QW
Quality-weighted signal pooling — noisy inputs reduce in influence, never break the state
Cluster sub-groups — emotional coalitions detected within larger participant sets

How signals become a collective state

Four aggregation methods, working simultaneously.

01
Quality-Weighted Pooling

Rather than treating all signals equally, the system weights each participant's contribution based on the integrity of their sensor data. Noisy channels reduce in influence — the aggregated state reflects what can be trusted, not simply what is present.

02
Three-Dimensional Emotional Mapping

Individual biometric signals are translated into coordinates across three axes: valence (pleasantness / discomfort), arousal (calm / activated), and coherence (fragmented / unified). Together, these three dimensions describe the emotional landscape of the group with nuance and precision.

03
Coherence Calculation

Cross-correlation and synchronization metrics quantify the degree to which participants' physiological states are aligning. High coherence signals moments of shared presence; low coherence reveals fragmentation or transition. This single measure is one of the system's most expressive outputs.

04
Sub-Group Cluster Analysis

In larger groups, emotional coalitions often form — subsets of participants whose physiological states align more closely with each other than with the whole. The system detects these clusters and preserves their dynamics within the collective state representation, enabling more nuanced sonic output.

"The system does not average its participants — it discovers what they share, preserving the texture of difference within the shape of the whole."
Emphonic System — Design Principle

Architecture

A tiered architecture ensures aggregation is always current.

The aggregation layer operates across two processing tiers simultaneously — each serving a different temporal demand. The immediate experiential loop runs at sub-100ms. Short-term pattern refinement operates within one second. Together they keep the collective state responsive and intelligent.

T1
< 20ms
Data Routing

Core signal acquisition and safety-critical data handling. Biometric flow is guaranteed at this tier.

T2
< 100ms
State Synthesis

Physiological signals are fused into the collective state. This is the primary aggregation tier — feeding the sound and visual system with a continuously updated portrait of the group.

T3
< 1s
Pattern Refinement

Short-term emotional trajectory analysis updates the aggregated model. Cluster boundaries shift. Coherence trends are smoothed. The collective state becomes richer.

T4
< 10s
Memory Encoding

Session-level aggregated states are encoded into long-term memory for the Evolution layer.

← 01 — Sensing 02 of 04 03 — Translation →