Methods for Accurate Acoustic Characterization with Ultra-Low Noise and Minimal Effect from Acoustic Reflections

Abstract

It is common to measure the response of devices and structures to sound due to an imposed sound source. Unfortunately, acoustic reflections from walls and/or instruments often contaminate the results. In this paper, methods of acoustic characterization are described to minimize the influence of acoustic reflections. It is shown that this process results in clean and smooth data. A simple time-domain window is implemented for diminishing the contribution from reflection waves. Moreover, a single frequency curve fitting approach is employed for better parameter identification and noise reduction, compared to traditional fast Fourier Transform analysis. Results obtained from a theoretical acoustic model with a reflection source are compared with measured results. Different cases of data acquisition with time and frequency analysis are experimentally demonstrated and validated. All experimental measurements are performed in an anechoic chamber. Results show that the approach presented here significantly reduces noise and also the influences of reflection waves in experimental data acquisition outcomes.