Development and evaluation of a digital MEMS microphone array for spatial audio

Abstract

We present the design of a digital microphone array comprised of MEMS microphones and evaluate its potential for spatial audio capturing and direction-of-arrival (DOA) estimation which is an essential part of encoding the soundscape. The device is a cheaper and more compact alternative to analog microphone arrays which require external-and usually expensive-analogto-digital converters and sound cards. However, the performance of such digital arrays for DOA estimation and spatial audio acquisition has not been investigated. In this work, the efficiency of the digital array for spatial audio is evaluated and compared to a typical analog microphone array of the same geometry. Our results indicate that our digital array achieves the same performance as its analog counterpart, thus offering a cheaper and easily deployable device, suitable for spatial audio applications.

Comparison between analog and digital phased microphone arrays for aeroacoustic measurements

“…Although many comparative studies have been conducted on the acoustic imaging algorithms themselves [13,[17][18][19][20], there is no extended literature evaluating the influence of the hardware used in a phased array. Alexandridis et al [16] obtained similar results when using an analog and a digital microphone array.…”

“…In this configuration, the cables connecting the microphones to the DAS should have good electromagnetic shielding properties, otherwise electromagnetic interference (EMI) will degrade the data, as the signal inside the cable is analog. In general, this results in a rather complicated and expensive system [16]. Recent technological advances in digital microphones, such as micro electromechanical systems (MEMS), offer a less expensive and simpler alternative with respect to conventional analog microphones for phased array applications.…”

“…Recent technological advances in digital microphones, such as micro electromechanical systems (MEMS), offer a less expensive and simpler alternative with respect to conventional analog microphones for phased array applications. This type of microphones is normally found in modern cellphones and are, therefore, usually optimized for the human speech frequency range (approximately from 300 Hz to 3 kHz) [16]. Digital sensors are more complex as they have the filtering and ADC embedded, see Fig.…”

“…Therefore, having this feature embedded in a cheap sensor can save a reasonable amount of the budget. On the other hand, it is difficult to ensure the precision of this data conversion [16]. Although many comparative studies have been conducted on the acoustic imaging algorithms themselves [13,[17][18][19][20], there is no extended literature evaluating the influence of the hardware used in a phased array.…”

Multiple Source Location Estimation on a Dataset of Real Recordings in a Wireless Acoustic Sensor Network

“…All the arrays had 8 microphones. (0, 200) 3 cm MEMS integrated in array [13] of 44.1 kHz, while the digital ones (A 1 and A 4 ) operated at 48 kHz. Thus, after recording, the signals from arrays A 1 and A 4 were downsampled to 44.1 kHz in order to keep the sampling frequency constant at all nodes in the WASN.…”