Page 19 - Acoustics Today

Fig. 1. A National Park Service (NPS) acoustical monitoring system deployed in Zion National Park. Through many generations of NPS systems, wind speed logging has been a consistent component, to identify times when pseudonoise generated by airflow around the wind screen would inflate the spectrum measurements.
system will be similar to a solid disk of the same size mounted in a baffle. For fre- quencies such that the trans- ducer diameter corresponds to multiple wavelengths, the horn microphone will be highly directional. In Table 1, a 20 mm mouth diameter was assumed, which implies highly directional response
12
toring of natural sounds and biological studies. Accordingly, many DAR features are superfluous– such as equalization or spe- cial effects—or detrimen- tal—such as a visual inter- face that continually draws power. Nonetheless, the core function is to record high fidelity acoustic sig- nals to memory and most DARs can be repurposed for unattended recording in harsh environments by enclosing them in a rugged weatherproof housing, providing a supplemental battery power supply, and shrouding the micro- phones with a wind fairing and splash protection.
Selecting a DAR involves consideration of several characteristics. In terms of cost and power consumption, the DAR will almost always be the most critical selection. The DAR will rarely be larger and heavier than the power subsystem, but the DAR power consumption largely determines the power requirements. Except for
applications that demand enormous data storage capacity, DARs that use solid state memory will be preferable, due to significant reductions in power consumption and greater tol- erance of environmental shocks relative to spinning hard disks. Intrinsic noise levels, dynamic range, bandwidth limits, sensitivity to environmental conditions, and build quality are other features of interest. The suitability of any DAR for acoustic measurements will depend upon the application and the environment. In some situations an extremely inexpen- sive unit may be the best choice, despite the limitations this imposes on the accuracy of the measurements.
The capabilities of consumer audio recorders to record accurate time histories of acoustic signals have been docu- mented.15 Compressed audio (MP3) can be calibrated and
16
and equivalent input noise (EIN).
Power consumption is fairly straightforward to compare.
The Centon would require less than 15 Amp-hours of battery to run for 30 days; it could run for a month on a single D size
The per- formance gain of a horn comes with a cost—nonuni- form gain across frequency. In general, horns deempha- size low frequency sounds, with the cutoff frequency for maximum gain being deter- mined by the flare rate of the horn. This low frequency de- emphasis can be advanta- geous for two reasons— background sound level spectra are often red,13 and high frequency sounds expe- rience greater atmospheric absorption. Horns can sup- ply more gain to extend high frequency detection capabil- ity while suppressing the ability of low frequency
at about 16.5 kHz.
noise to saturate the system. Examples of the theo- retical on-axis frequency
response for two exponential and conical flared horns are shown in Fig. 1. The dimensions are as for the first example in Table 1 (i.e., a throat diameter of 0.25 mm and mouth diameter of 20 mm) and the length is varied to achieve vary- ing cutoff frequencies. The gain of the conical shaped horn approaches the maximum at a more gradual rate than the exponential. The marked peaks and troughs in the response functions can be minimized by adjusting the dimensions of the horn such that the mouth impedance is matched to the
14
radiation impedance. As a result, a smooth frequency
response is achieved, e.g. the exponential horn in Fig. 1 with fc = 5.5 kHz.
Recorders
The increasing capabilities of consumer digital audio recorders (DAR), especially increases in storage capacity and reductions in power consumption, enable continuous audio recordings exceeding one month in duration with packages that are relatively small and inexpensive. These devices uti- lize integrated circuits for signal conditioning, analog-to-dig- ital conversion, and data storage, reducing the cost and com- plexity of assembly. Most DARs are designed for the purpose of speech or music recording rather than unattended moni-
There are many consumer recorders to choose among, and capabil- ities are evolving rapidly as technology progresses. Table 2 displays specifications for a representative set of DARs including power draw, storage capacity given memory type,
18 Acoustics Today, July 2012
used to calculate sound level metrics from audio data.