Optical methods for modeling and experimental detection of sound sources behind windows

Abstract

Detecting sound from a distance is not a novel concept and laser vibrometers are often used for this purpose. A potentially better method is to measure the reflection of the usually stationary environment in a vibrating window. Experiments were conducted to capture high-speed video of vibrating windows with sound sources located inside a building. Low frequency vibrations produced by a speaker operating between 19 and 37 Hz were clearly visible on high contrast reflections under a very limited set of conditions. To better understand the limitations of the method in terms of variabilities of potential scenarios with sound, window, optical reflection, camera and analysis methods a parallel modeling effort was initiated. The detection of sound using optical methods can vary with sound source parameters (e.g. frequency, spectrum, amplitude, …), the window structure (glass thickness, number of panes, elasticity,…), the character of the reflected environment (contrast ratio, intensity,…), the measuring equipment (camera frame rate, spatial resolution, dynamic range, integration time, noise level,…), and the analysis methods (short term Fourier analysis window size, apodization, preprocessing steps,…). This talk will highlight some of the findings to determine the feasibility of optical detection of sound sources behind windows.