## Overview
Acoustic surveillance arrays are distributed networks of directional microphones and acoustic processing hardware deployed across urban environments to capture, classify, and analyze sound at scale. In GLMZ, acoustic infrastructure exists in three distinct tiers: the municipal SoundGrid operated by the Civic Security Authority, proprietary corponation acoustic networks within sovereign zones, and the emergent mesh of always-on audio capture embedded in consumer devices, smart fixtures, and public information terminals. These three layers are not formally integrated but share classification models and, under certain law enforcement access protocols, data. The combined effect is near-total acoustic coverage of any space that contains people and technology.

## Capabilities: Classification and Content
Acoustic arrays perform two fundamentally different types of analysis. Classification-level analysis identifies sound events — gunshots, breaking glass, crowd disturbance signatures, vehicle collision profiles — and triggers automated dispatch or alert protocols. This function is widely accepted as legitimate and is politically uncontroversial. Content-level analysis — the capture and processing of speech — is legally restricted but technically unrestricted. Current beamforming algorithms can isolate individual conversations in crowd environments at distances of up to 25 meters with intelligibility rates sufficient for automated transcription. The SoundGrid's official operational doctrine states that content capture requires a judicial proxy authorization, but the system's architecture makes it technically possible to retroactively retrieve buffered audio for any location within a 6-hour window, subject only to an after-the-fact logging requirement.

## Voice Biometrics and Identity Linking
Acoustic arrays are increasingly integrated with voice biometric systems that can match captured speech against the Civic Voice Register — a database of voice samples collected during civic registration processes, indenture contract signings, and voluntary enrollment. A confirmed voice biometric match links acoustic capture data to a specific civic identity record, enabling the reconstruction of a subject's speech history across public and semi-public spaces. This capability is not widely publicized, and the Civic Voice Register's existence was not officially acknowledged until a data disclosure lawsuit in 2194 forced partial transparency. The register currently contains confirmed voice profiles for approximately 67% of GLMZ's indexed population.

## Counter-Acoustic Technology
The counter-acoustic device market has grown significantly in parallel with array deployment. Personal acoustic masking devices — small wearable emitters that generate a locally canceling noise field around a speaker — are available commercially and are not themselves illegal, though their use in certain contexts (proximate to law enforcement operations, within corponation zones) may trigger secondary surveillance attention. More radical solutions include subvocalization implants that allow speech to be transmitted to a recipient device without producing audible sound, effectively moving conversation below the threshold of acoustic capture. These implants are expensive, require neural interface infrastructure, and introduce their own surveillance vulnerabilities at the neural data layer.