Comparison of Three Linear Digital Filters Applied to Improve the Quality of the Measured Acoustic Field

Abstract

The measurement of acoustic fields generated by ultrasonic transducers is important for determining the focal length, lateral resolution, and amplitudes of the lateral and grating lobes. The acoustic field is commonly characterized by a set of scans using a needle hydrophone. The output of the hydrophone can be connected to an analog filter to enhance the signal. However, the analog filter might not be sufficient to avoid the noises that distort the signals. Alternatively, linear digital filters can be advantageous to improving the acoustic-field characterization. In this work, three filters were investigated: moving average (MA), band-pass Hamming window (HW), and band-pass Blackman window (BW). The filters were implemented and evaluated in terms of the root-mean-square error (RMSE) of the measured sound field, which was filtered, in relation to the simulated acoustic field (gold standard). As a compromise between effective filtering and signal non-distortion, a method to model the MA kernel length was proposed. All the filters reduced the noise of the measured acoustic field. The HW and the BW filters were more effective (RMSE = 4.01%) than the MA filter (RMSE = 4.28%). In spite of the small quantitative difference, acoustic field comparisons showed qualitative improvements.