AI-Driven Soundscape Analysis for Citizen Science

PlatformMobile & Web
Client App Websitehttps://apps.apple.com/us/app/soundmaster-rva/id1291725677
AreaSoundscapes, Sound Analysis, Acoustics

AI-Driven Soundscape Analysis

HARMONY is a prototype platform, aiming to provide a multimodal set of features, functionalities and options to consortia, organizations and individuals around the fields of sound engineering, audio recording, geolocalization and soundscape.

The platform’s versatile structure, stack and configuration provide opportunities for use in education, audio engineering, communication & media as well as medical contexts.

The platform consists of a mobile client app, performing from simple to highly advanced sound processing and recording and a server application (still at a prototype level), which can be used to facilitate storing of audio recordings, location and audio metadata as well as orchestrate and facilitate learning scenarios in the form of linear and non-linear treasure hunt games.

The client app provides (but isn’t limited to) the following technical features and functionalities:

• Main Features
– RT60 measurements
– Octave-based audio analysis
– Professional sound level metering
– Advanced calibration and equalization support

• Under the hood
– 44.1kHz/32bit raw audio recording
– 20Hz to 20kHz frequency response
– 90dB dynamic range, from 25dB to 115dB SPL[A]
– Low latency and processing-loss spectral analyzer

• Reverberation Time Measurements
– Octave-based RT60 measurement capability
– Interrupted noise method implementation through signal excitiation
– Build-in band-pass input filtering for improved tolerance to background noise
– Build-in band-pass noise generator
– Option for setting the repetition-count for each measurement
– Ability to detect measurement errors and inform the user through an error-code protocol
– High background noise, insufficient signal volume, low SNR and noise during decay errors detection
– Retry on error automation

• Noise Measurements
– Digital VU and numeric visualizations
– Broadband RMS, Peak, Max, Min, Leq and measurement duration data handling
– Octave-based RMS, Leq, Max, Min data handling
– Impulsive, Fast and Slow response modes
– A, B, C, D, Z weighting filters

• Advanced Calibration
– Calibrated and equalized for each iPhone model
– Calibration data is OTA (Over the Air) updated
– Additional user calibration and equalization banks for both embedded microphone and external input source
– Custom octave-based gain adjustment support

• Session Management
– Overall statistics calculation for noise measurements
– Photo capture capability
– Time and geolocation awareness support
– Title, location and comments metadata support
– Audio recording (CAF format), measurement data (CSV format) and metadata export capabilities

• Additional Capabilities
– Option for keeping the device awake during recording
– Option for auto-stopping measurements after a pre-configured duration
– Background recording capability