Smartphones are ubiquitous, powerful devices, capable of running a wide range of apps that make use of the various sensors installed in the devices. One such sensor is the microphone – traditionally used for phones calls – which can also be adapted for use in a wide range of other apps. These microphones are relatively inexpensive, mass-produced devices, with varying acoustic properties. The manufacturers of smartphones do not freely publish technical information on the properties of their microphones.
In certain apps, e.g., sound level meter app, restaurant or nightclub noise rating app, etc., it is important to know how the microphone in a particular phone model behaves in terms of its sound pressure response and frequency response. Some apps include a feature which enables the phone’s sound level to be calibrated against a reference measurement taken on a higher precision sound level meter. One problem with this calibration approach is that most people do not have access to a high precision sound level meter, and the uncalibrated reading on the phone is likely to be quite erroneous.
The aim of this project is to develop a robust calibration procedure and test rig for smartphones to improve their accuracy when used as sound level meters or similar in apps (where a degree of engineering accuracy is required from the microphone). This calibration procedure will be based on the assumption that models of phone from a particular manufacturer (e.g., Samsung S22) will all use the same microphone and manufacturing tolerances, so that a single set of calibration factors can be determined and applied for any of that model of phone. Calibration data would include a correction factor for the measured dBA sound pressure level and a spectral correction for phone microphones which do not have a flat frequency response.The outcome of this calibration procedure would be to use it to develop a calibration database for different models of makes of phone.
Key features of this project will include:
• Digital signal processing (Matlab) • Physical construction of a test rig • Analysis of measured acoustical dataProject outcomes
• Produce a mobile phone sound level meter calibration rig and test procedure • Develop signal processing code (Matlab) to determine the calibration factors for different phones • Create a database template for different models and makes of phone
Confidentiality
Not confidential.
Acoustical Society of New Zealand Best Student Project Prize: $500
This project is eligible to be judged for a prize of $500 NZD from the Acoustical Society of New Zealand for the best student project. Judging will occur during the project presentation at the end of the year.
The Acoustical Society of New Zealand has a particular interest in the outcomes of this app for use in conjunction with the SoundPrint venue noise rating app: www.soundprint.co/
Industry Sponsorship
This project will receive industry sponsorship from Norman Disney & Young.
Names of co-supervisor and assessor:
• Andrew Hall – University of Auckland • Tim Beresford – Norman Disney & YoungUndergraduate
None
Acoustics Lab (City 422.154, Lab)