QuestionandAnswers1.pdf

SBIR 2015-1

Questions and Answers #1 November 5, 2014

Q: On Subtopic 8.2.1F: Given that Phase I is asking for both modeling as well as development and initial testing of a prototype, all within a Phase I budget, we are wondering whether or not NOAA would be providing any of the sensors as well as their telemetry packages as Government Furnished Equipment (GFE)? If "yes", what are these sensors, including their specifications, and how many would be provided?

A: We have developed a hydrophone/pre-amp combination that works extremely well together for our purposes. We can provide the pre-amps and full design specifications for these pre-amp (GFE). We will not, however, provide the hydrophones (HTI 96-min, which can be purchased for approximately $350/each, see http://www.hightechincusa.com/96_MIN.html). Other hydrophone/pre-amp combinations may be considered; however, design/building/testing will not be provided by the Government.

Q: On Subtopic 8.2.1F: What are the expected performance specifications for the resulting prototype Nonlinear Towed Hydrophone Array?

A: The nonlinear towed hydrophone array prototype will be expected to be robust and stable, as well as have low self-noise and low flow-noise when towed under NMFS standard operating procedures for shipboard cetacean surveys. Robust refers to the ability of the array to be towed 300m behind ship at 10 knots for months on end. The device must be stable: it should have little or no rotation when towed, and this rotation should be quantified in real-time (we have had success with the Adafruit ADXL335 accelerometer). Nonlinear arrays, by nature, have increased self-noise and flow-noise. This noise must be minimized such that detection of biological sounds between 1-180 kHz can be detected during standard operating procedures. Our experimental designs to date have made great strides in identifying and mitigating this noise, we are happy to share these details with any company awarded this SBIR. I have provided a recent report that summarizes recent designs and testing (Supplemental Information #1)

Q: On Subtopic 8.2.1F: Will NOAA provide any testing support to the Phase I contractor? If "yes", what will the testing support include and where?

A:NOAA can provide some degree of field testing (at sea) out of San Diego. NOAA personnel at Southwest Fisheries Science Center in La Jolla can provide a minimum of 2 days field testing during Phase 1 out of San Diego. Field tests would be conducted on a local fishing vessel (approximately 40'- 65') in offshore waters. SWFSC can provide 12v marine batteries, acoustic recording hardware (including computer) and 2 scientists to aid in the testing. We can provide a limited deck cable for testing (maximum 8 paired wires). In our experience this will provide 4 hydrophones, 1 depth sensor, 1 accelerometer, 1 seawater ground, and 1 power supply. Testing with any SWFSC cable requires mating connectors (Teledyne Impulse MHDM-26).

Q: On Subtopic 8.2.1F: What type of vessel will employ the Nonlinear Towed Hydrophone Array and what are the vessel's performance characteristics that the nonlinear array might be required to meet (speed, max sea state, data rate, speed of data computation and display, mode of data presentation, etc)?

A:Vessel characteristics vary, but typically we expect survey speed of 10 knots and a maximum sea state of Beaufort 6. Actual swell height will vary, but typically we work in swell height less than 12 feet. We have included documentation outlining a version of our Acoustic Recording System (Supplemental Information #2), which has been modified to include a National Instruments 6356 http://www.hightechincusa.com/96_MIN.html with a sampling rate of 500kHz. Depth and accelerometer data are digitized using the MCDaq USB- 1608G. Digitized audio, depth, and accelerometer data are input into Pamguard software (with real-time GPS data). This software is open-source and available at pamguard.org. Researchers at NOAA's Southwest Fisheries Science Center will provide these hardware/software components for field testing and limited lab-testing.

Q: On Subtopic 8.2.1F: If this project progresses to Phase II, what quantities of Nonlinear Towed Hydrophone Arrays are anticipated?

A:This will depend on cost, funding, and expected participation in surveys. Beyond NOAA, there is increasing demand for off-the-shelf towed hydrophone arrays for use in private/institutional research and mitigation surveys. We expect the need to increase in the coming decade. If Phase I identifies a successful prototype, we can aid in identifying and quantifying this need.

Q: On Subtopic 8.2.1F: If Phase II is successful, does NOAA anticipate the Nonlinear Towed Hydrophone Arrays moving into a Phase III program? If "yes", what quantities of Nonlinear Towed Hydrophone Arrays might be anticipated during Phase III?

A: (see above) This will depend on cost, funding, and expected participation in surveys. Beyond NOAA, there is increasing demand for off-the-shelf towed hydrophone arrays for use in private/institutional research and mitigation surveys. We expect the need to increase in the coming decade. If Phase I identifies a successful prototype, we can aid in identifying and quantifying this need.

Please note that the working group names were removed from the following documents. All questions in regards to this subtopic must be forwarded to the Contract Specialist (Joan Clarkston). In order to ensure fair competition, all communication must be routed through the CO to ensure all potential offerors receive the same information.

http://pamguard.org/ http://pamguard.org/

NOAA Advanced Sampling Technology Working Group

FY2013 Final Report to the Office of Science and Technology – October 2014

TITLE

Development of a hydrodynamic towed tetrahedral hydrophone array to improve localization accuracy during shipboard line-transect cetacean surveys

GOALS

Passive acoustics has been increasingly used for detection of cetaceans during shipboard surveys conducted by all NMFS Science Centers. Towed linear arrays are currently implemented into line-transect cetacean abundance surveys, but are subject to many limitations. Tetrahedral arrays (or other special arrays) have been shown to address some of these limitations, but designs that have been tested to date were limited to tow speeds of less than 6 kts. In order to operationalize the tetrahedral array into standard NOAA shipboard cetacean surveys, the design must be modified for operation at 10 knots from NOAA research vessels. A streamlined, hydrodynamic tetrahedral array will improve the tracking of animal groups, decrease the time necessary to assess each group, and allow more accurate data collection overall. Implementing the tetrahedral array into the modular array systems of each Science Center may provide a more efficient and precise method of acoustic data collection to improve the accuracy of cetacean abundance estimates.

Specifically, we proposed to:

8.2.1F Supplemental Information #1 - Final Report

(1) research and develop a streamlined, hydrodynamic design for towing the tetrahedral array at faster speeds on large ships,

(2) test and assess the performance of tetrahedral array prototypes to evaluate necessary modifications, and

(3) produce operational tetrahedral array prototypes to be used during PIFSC and SWFSC cetacean cruises, and available to loan to other Centers to test as part of their survey efforts.

PRIORITIES

The Marine Mammal Protection Act mandates the assessment of all marine mammal stocks within U.S. Exclusive Economic Zones (EEZ). NMFS has met this mandate through use of visual line-transect surveys since the 1980s. Passive acoustic monitoring methods were implemented in the form of towed hydrophone arrays in the 1990s and have since evolved to provide assessments of the number of groups missed by the visual survey team, have allowed for acoustic tracking of vocal groups, and for some species, have provided independent estimates of abundance based on the acoustic data (Barlow and Taylor 2005).

While linear towed hydrophone arrays are excellent tools to detect cetaceans during

NMFS population surveys, it is rarely possible to determine the location of a vocal group when they are first detected (Rankin et al., 2008b). A distance estimate from a linear towed array requires the convergence of successive bearing angles to vocalizations over time as the animals pass the beam of the ship. However, the time necessary to obtain bearings to animals passing the beam can result in the animals becoming quiet, responding to the vessel, or other scenarios that create uncertainties in the data, and may introduce bias into the location estimates. Linear arrays are also subject to a left/right ambiguity that cannot be resolved without turning the vessel, which interrupts the effort of the visual observer team.

Previous field tests on a smaller vessel proved the effectiveness of a tetrahedral array for three-dimensional localization at speeds up to 6 kts. This array can discriminate on which side of the ship the vocal group is located and provide the approximate depth of the animals depending on their distance from the ship (Southall et al., 2012). Although the tetrahedral array alone cannot provide instantaneous distances to a vocal group, it can provide more precise bearing information and when combined with a linear array should provide more precise location estimates with fewer bearings.

Our priority with this project is to redesign the tetrahedral array to improve the hydrodynamics to allow for towing at faster speeds from large NOAA ships. A number of conceptual designs were evaluated and the two with the lowest modeled drag and flow noise were constructed for field testing.

APPROACH

A tetrahedral or spatial array is hydrophone array with four elements that are not in the same plane. The main benefits of the tetrahedral array include its ability to discriminate the side of the ship that the animals are located and their relative depth, which will allow us to track animals using a more direct route, collect more precise perpendicular distance estimates to animal groups, and collect acoustic data more efficiently overall. We collaborated with an expert team of marine engineers, led by Gerald D’Spain at the Scripps Marine Physical Laboratory

(MPL), to incorporate our specific requirements into the array design. Sulaman Ahmed, a UC

San Diego engineering student and intern for the MPL, drew and tested two tetrahedral designs using CAD modeling software under the guidance of Dr. D’Spain and marine engineers, David

Price and Scott Jenkins. The main design features included a rigid, hydrodynamic structure that maintained a specific three-dimensional configuration of hydrophone elements necessary for precise 3-D localization. The structural frame was designed to be a manageable size for one person to handle during deployment and retrieval. Other design considerations included maximizing the distance between hydrophones (within the manageable size constraint), lightweight structural materials, and creating contours that minimized flow noise. In addition, current preamp designs should be adequate to reduce ship and flow noise to similar levels experienced with existing linear arrays. We also included an accelerometer as a new internal component to measure the rotational movement of the array. These design elements were intended to make the arrays suitable for towing at 10 knots. Both designs were tested during several sea trials to determine the necessary modifications and assess the performance of each prototype, including their stability and flow noise. Once we created a fully-functional prototype, it would be made available to other Science Centers to test on their NOAA shipboard cetacean surveys to gain feedback and eventually operationalize the tetrahedral array throughout each region.

WORK COMPLETED AND RESULTS

High-Speed, Hydrodynamic Array Design

In order to accurately draw and model the design of the array, marine engineers used

SolidWorks 3D CAD modeling software (Dassault Systemes, 2010a), allowing us to visualize the angles, lengths, shapes, and overall size of the array (Fig 1). The resulting shape was intended to create a more hydrodynamic structure for towing at faster speeds. Our requirements for the frequency range of signals to be detected also contributed to the dimensions of the array design. Other details, such as avoiding entanglement in kelp and having the array be a manageable size for one person to handle, were also taken into account in the design phase.

Once drawings in Solidworks were finalized, the designs were tested using SolidWorks

Flow Simulation software (Dassault Systemes, 2010b). This intuitive computational fluid dynamics (CFD) tool calculated the spatial distribution of the flow velocities and hydrodynamic pressures associated with the array immersed in a steady-state background fluid flow. We tested two array designs using a velocity of 5 m/s (10 knots) to simulate the flow of water over the array. We also changed the direction of the flow over the array at various angles, including 0, 1, 2, 4, and 8 degrees, to test how the turbulence changes with the different “angles-of-attack.”

With these different tests, we were able to evaluate whether or not the drag on the array will be sufficient to keep the array straight during towing at 10 knots, determine the torques due to the hydrodynamic forces acting on the array and if the design dimensions keep the array from twisting, and to obtain a picture of where flow turbulence is likely generated and therefore where flow noise may occur over the hydrophones.

Results of the simulation tests provided information on what aspects of the design worked well and what aspects needed modification. After incorporating the modifications into the SolidWorks drawings, the flow simulations were repeated. This entire process was repeated enough times to find the optimum dimensions to minimize the turbulent flow over the array while allowing for adequate spatial resolution of the hydrophones to detect and localize broadband signals from different cetacean species.

Figure 2. The initial stages of the flooded torpedo tetrahedral array prototype (left) show the separate, hydrodynamic structural shell that houses the electronic components. The complete, modified version (right) displays the weighted, tri-fin setup for improved stability.

Figure 1. SolidWorks drawings of two tetrahedral array designs, the X-array (left) and the flooded torpedo array (right) that were considered.

Figure 3. Final CAD drawing of the X-array design (left), with hydrophone pods (A), struts, (B), and stream-lined array body (C). The fully-assembled X-array (right) displays the larger spacing between hydrophones.

Sea Trial Results

Construction of the flooded torpedo and X-array prototypes (Fig 2, Fig 3) was completed in March and April 2014. The prototypes were tested during four sea trials between March and

August 2014 aboard dive vessels and NOAA ships. Due to the different times each array was completed, the modifications that were made, and other logistical constraints, only one sea trial tested both prototypes with the traditional oil array. The sea trials focused on assessing the flow noise levels, stability, and localization accuracy for each array prototype.

Flow noise levels were examined for both arrays using recordings from the single sea trial where all arrays were tested. This sea trial took place aboard a chartered dive vessel after the first modifications were made to each prototype. A recording from each array was examined during tow speeds of 5 knots, a speed that provided a quieter baseline and maximum stability for a better comparison of each array. The noise spectra of one hydrophone channel from each array were compared (Fig 4). We found that the oil array and torpedo prototype had very similar noise levels overall, while the X-array was significantly noisier at all frequencies up to 10 kHz. We attributed the flow noise to the design of the struts and hydrophone pods (Fig 3a,b), which allowed sea water to flow directly against the hydrophone cables and sensors. After several tests, we potted the hydrophones in solid polyurethane to decrease the noise to more reasonable levels that would not impede the detection of animals. Further tests for the X-array took place aboard a

NOAA ship with a different recording system, which could not be directly compared to previous noise measurements. However, during these tests the X-array hydrophones were very quiet, similar to noise levels of an oil-filled linear array. We have not yet been able to perform sea trials

A

B

C to test the “potted” X-array with the torpedo prototype and oil array for a second comparison, though this is planned for the later portions of the CalCurCEAS.

Figure 4 Noise spectra for three array types during a sea trial. The X-array is relatively louder at lower frequencies below 1 kHz.

Both array prototypes contain an internal accelerometer to measure the rotational movement while towed at different speeds. Initial stability tests of the torpedo array showed it to be rapidly rotating at approximately 120 degrees every second (Fig 5a). The unstable orientation did not appear to introduce noise, but greatly inhibited any attempt to test its localization capabilities. The torpedo array’s stability was addressed with a modification, a weighted tri-fin system attached to the aft end of the array body. The rotational movement drastically decreased allowing it to be towed at speeds from 8-9.5 knots on smaller research vessels (Fig 5b). The localization accuracy was also assessed from a smaller vessel by comparing real-time directional bearings from the array to sighted animal groups. Localizations proved to be accurate within several hundred meters.

The stability of the X-array proved to be a more complicated issue. During its initial sea trial aboard a NOAA ship, the streaming accelerometer data indicated it to be fairly stable (Fig

6a). However, once the hydrophones were fully potted to correct the noise issues, the X-array began to rotate 360 degrees when tested a third time from a NOAA ship (Fig 6b). It is unknown whether the initial accelerometer data was incorrect or misinterpreted, or if the modifications affected the balance of the X-array. Several attempts to improve the stability were made, such as adding weight to the bottom strut, but further testing is required. Due to the stability issues, the localization accuracy of the X-array was not assessed.

Figure 5 Graph of the accelerometer data for the first sea trial of the flooded torpedo array prototype (A) shows the array rotating rapidly, about 120 degrees per second. After modifications, the prototype displayed a stable orientation with less rotational movement of about 20 degrees (B).

Figure 6 A plot of the raw accelerometer data shows a stable orientation of the X-array hydrophones (A). Once the

X-array was modified, the orientation became less stable causing the array to spin (B). The blue rectangle represents approximately 1 minute of missing data.

IMPACT/APPLICATIONS

There are several different operational scenarios within which cetacean assessment operations may be improved with the incorporation of a tetrahedral array design. By using a tetrahedral array alone, we can determine acoustic bearings to animal groups from the ship without the left/right ambiguity, saving significant time when guiding the visual team to the location for species confirmation. More efficient acoustic tracking methods may also improve the enumeration of animal groups that become more widely dispersed throughout the length of

A B

A B s the encounter. It is also possible to obtain the relative depth of individual animals or groups, which may increase the probability of sampling or tagging deep-diving, cryptic species. Overall, these are significant improvements in acoustic detection applications that will provide more information about the different cetacean species in each region.

TRANSITION TO OPERATIONS

Based on the results of the sea trials and data analyses, both tetrahedral prototype designs require further modifications and testing before implementing them into NMFS shipboard cetacean surveys. The flooded torpedo prototype is quiet and stable up to 9.5 knots from smaller research vessels and will require further testing to make it operational aboard a NOAA ship. We plan to test the torpedo prototype from a NOAA ship for the first time during the next leg of the

SWFSC cetacean survey (CalCurCEAS). However, the small spacing between the tetrad of hydrophones (~3.5 cm) makes the flooded torpedo array better-suited for localizing echolocation clicks instead of whistles. The X-array design includes a larger tetrad (~60 cm) to accommodate the longer wavelengths of whistles, theoretically detecting animals from greater distances. The stability of the X-array prototype is the main issue still needing to be addressed. Additional modifications are being made to its design in order to test it on CalCurCEAS as well.

In order to further the development of either prototype, we submitted a proposal to the

Small Business Innovation Research program (SBIR). This is an attempt to allow the private sector the opportunity to resolve the noise and stability issues using the existing designs.

Once a fully-functional prototype is developed, we plan to make both arrays available to other Science Centers to use and test during their cetacean surveys. Each region focuses on different cetacean species and different assessment goals so it will be useful to gain feedback from other NMFS programs to help improve the tetrahedral array for each Science Center. Once the tetrahedral design has been thoroughly tested, our intention is to seek additional funds to organize a workshop similar to Rankin et al. (2011) in order to completely operationalize the tetrahedral array throughout each NMFS region. These arrays would be built and tested to optimize acoustic data collection based on needs of each Center.

PUBLICATIONS AND PRESENTATIONS

None at this time.

EXPENDITURES

The funds received for this project were spent on a various items related to the design and construction of the tetrahedral array. Table 1 displays a cost breakdown of these items. Funds totaling $10,866 were set aside for a software modification to accommodate an incoming stream of accelerometer data to adjust the localization output. These funds were not allocated to our vendor in time within FY2013 due to a number of problems within the contract award system.

These funds were spent by the PIFSC Director’s Office and will be replaced in FY14 to allow for these required software modifications.

Item Cost

Engineer contract $18,000

Y. Barkley salary $45,000

Vessel charter $15,000 tetrahedral array supplies $8,600

Total $86,600

LITERATURE CITED

Barlow, J., and B. Taylor. (2005). Estimates of sperm whale abundance in the northeastern temperate Pacific from a combine acoustic and visual survey. Marine Mammal Science.

21(3), 429-445.

Dassault Systemes. (2010a). Driving better product design with SolidWorks simulation.

SolidWorks Corp, white paper,12.

Dassault Systemes. (2010b). The role of CFD in real life designs. SolidWorks Corp, white paper,12.

Rankin, S., Barlow, J., Oswald, J. (2008a). An assessment of the accuracy and precision of localization of a stationary sound source using a two-element towed hydrophone array.

NOAA Technical Memorandum, NOAA-TM-NMFS-SWFSC-416, 29.

Rankin, S., Barlow, J., Oswald, J., Ballance, L. (2008b). Acoustics studies of marine mammals during seven years of combined visual and acoustic line-transect surveys for cetaceans in the

Eastern and Central Pacific Ocean. NOAA Technical Memorandum, NOAA-TM-NMFS-

SWFSC-429, 58.

Rankin, S., Barlow, J., VanParijs, S., Soldevilla, M., & Oleson, E. (2011). Proposal for a workshop to develop standardized methods for integrated passive acoustic data collection methods during shipboard cetacean surveys. Advanced Sampling Technology Working

Group Proposal.

Southall, B., Calambokidis, J., Tyack, P., Moretti, D., Friedlaender, A., DeRuiter, S., Goldbogen, J., Falcone, E., Schorr, G., Douglas, A., Stimpert, A., Hildebrand, J., Kyburg, C., Carlson, R., Yack, T., & Barlow, J. (2012). Biological and Behavioral Response Studies of Marine

Mammals in Southern California, 2011 (“SOCAL-11”). Final Project Report.

Acoustic Recording System: A portable hardware system for shipboard passive acoustic monitoring of cetaceans using a towed hydrophone array

Technological advances in both software and hardware have lead to dramatic improvements in systems used for passive acoustic monitoring (PAM) of cetaceans from a variety of platforms.

While these advances have vastly improved the quality and quantity of data, the rapid evolution of hardware and software have made it difficult to allow for standardization and collaboration between researchers. Recently, government scientists responsible for PAM of cetaceans have been coordinating efforts in an attempt to standardize methods, including hardware and software, to allow for collaboration and comparison of results. In addition, standardization of methods will allow the various organizations to make the most of limited resources. One component of this standardization was to create a portable acoustic recording system that would be the primary hardware package for standard shipboard surveys using towed hydrophone systems and would allow recordings from the full range of cetacean sounds (up to 160 kHz).

Specifically, the objective of this project was to create a simple, compact Acoustic Recording

System (ARS) that would include:

- A rugged, portable rack-mounted case

- Access to ‘clean’ 12v power; switchable to 12v battery power.

- Amplifier with high-pass filter, multiple outputs (including headphone)

- Audio interface for mid- and high-frequency recording options

- Low-power computer and monitor for digital recording

- GPS

To this end, in 2010 Southwest Fisheries Science Center designed and tested the ‘ARS’ during a shipboard survey of the Hawaiian Islands. This report includes hardware specifications for the original design, problems encountered during the field testing, and suggested design for future

ARS. Our goal is that this ARS system will be adopted by most (if not all) NOAA science centers for their standard shipboard towed hydrophone systems; Pacific Islands Fisheries Science

Center (PIFSC) has built their first ARS following our suggestions.

Original ARS Design

Three Acoustic Recording Systems (ARS) were built with the same basic design and some differences in certain hardware configurations. All the hardware devices, including make and model, are given in Table 1. A bandpass filter was included on ARS-3 to provide an option for an anti-alias filter; there was no mid-frequency (192 kHz) audio interface on this ARS. The

ARS-2 included two mid-frequency audio interfaces (Motu Traveller). All the components

(except the transformer) were selected to run on 12v DC power.

8.2.1F Supplemental Information #2

Four components were not designed to be rack-mounted (computer, GPS, fuseblock, and NI-

DAQ). These components were mounted on a ½” plastic base (marineboard) which was mounted on full-extension drawer slides for easier access.

Table 1. List of components for ARS Design. Not all components were used on all ARS.

Figure 1. Front (a) and back (b) view of ARS-3, including (from top to bottom) Avens bandpass filter, Astron power supply, Magrec amplifier, Ni-DAQ audio interface, Garmin GPS, and

Logisys computer.

Figure 2. Front (a) and back (b) view of ARS-2, including (from top to bottom) Astron power supply, Magrec amplifier, two Motu Traveller audio interfaces, Ni-DAQ audio interface, Garmin

GPS, and Logisys computer.

ARS Field Test

Three variations of the ARS were built and field tested during a shipboard survey in 2010. A brief explanation of problems encountered and suggestions for future ARS designs are provided for each hardware component of the ARS. There were no problems with the rack-mounted box, GPS, or fuse box; these are not discussed in this section. A suggested design for future ARS is provided in the Appendix, including a complete list of hardware components and a few particulars regarding building the units.

Power Supply: One of the goals of the ARS was to provide clean 12v power supply to the hardware to minimize electrical noise associated with the ships’ power supply. The Astron linear power supply was found to be sufficient on the R/V McArthur, where noise associated with the ships power supply were less problematic than on the R/V Oscar Elton Sette. The noise associated with ships power on the Sette was too severe to be used at any point. It is suggested that the noise level associated with the ships power be assessed for each vessel, and addressed accordingly. Some vessels may work well with a 12v power transformer such as the Astron, some vessels may require completely independent 12v power supply using clean battery power.

Amplifier: The Magrec HP27 is a custom-built stereo amplifier that allows a high-pass filter with an adjustable gain and multiple outputs. The Magrec also provided input for a depth sensor.

An early version of the Magrec (Magrec-1) included a single option for high-pass filter (20kHz) to maximize the singal-to-noise ratio (SNR) for detection of high-frequency signals from beaked whales and porpoise. We found that this limited our options, and Magrec-2 and -3 were designed to allow for variable high-pass filtering and various gain control for each channel. We

Available from Ecologic Ltd, U.K. http://ecologicuk.co.uk/Equipment.htm found that three independent outputs per channel was sufficient; however, we suggest that future designs include independent gain control for each channel.

Audio Interface (Mid-Frequency 192 kHz): A Motu Traveller was used as an audio-interface for mid-frequency sampling at 192 kHz. The benefits of the Motu are that it provides an internal low-pass filter set at the Nyquist frequency. This allows for recording at the selected sampling rate without additional anti-aliasing filters (which can be expensive). Significant problems were encountered with the Motu including: problematic software setup, issues with the computer recognizing the Motu which required frequent reboot or re-installation, and occasional unexplained automatic changes to the gain and sampling rate. Any one of these problems would be unacceptable, the combination of the three requires us to examine alternatives. Northeast

Fisheries Science Center (NEFSC) has had excellent success using ‘Fireface’ fire-wire-based audio interface. This has been field-tested during their east-coast surveys, and no problems were encountered. Fireface does not have an internal anti-aliasing filtration; however, this could be rectified by either including a high-quality filter such as the Avens, or by sampling at a rate that is greater than the frequency response of the hydrophones. Also, Fireface has a USB/Firewire version (UFX), a USB-only version (UC), and a 2-channel USB version (‘Babyface’) of their audio interface, . We have not yet tested any of these USB interfaces, but USB interfaces might be more robust than Firewire in some versions of Windows.

Audio Interface (High-Frequency >192 kHz): The National Instruments Data Acquisition

Board (NI-DAQ) acts as a sound-card with multiple channels and specific selection of sampling rate. These components have been used extensively in the field, with few problems. NI-DAQ boards do not have a filtering system; therefore, an anti-aliasing filter must be used unless sampling is greater than the frequency response of the hydrophone. Also, we have noted that there can be some electrical noise ‘leaking’ from uncapped terminals. We suggest purchasing and applying terminal caps to all unused terminals on the NI-DAQ.

Bandpass Filter: The Avens bandpass filter was included in one of the ARS, to provide an option for filtering at the Nyquist frequency to avoid aliasing. ARS that used the Motu did not need this filtering (due to the internal anti-aliasing filters), and the NI-DAQ sampled at 500 kHz, providing ample sampling above the frequency response of the hydrophones. In our experience, only high-quality filters should be used for anti-alias filtering, and high quality filters are prohibitively expensive. The high-quality Avens filter used here had been in our inventory for years, but we have found that the technology has improved to the point where we are able to sample at very high frequencies (500 kHz) and record to digital media without great expense. We suggest that these methods would most likely be preferable to investing in high-quality expensive filtering hardware.

Conclusions

The ARS designed and built by SWFSC allowed for relatively easy shipping overseas as well as shipboard setup. A 12v powered system decreased internal noise associated with ships’ power, although some vessels with extremely noisy ships power may need to completely isolate their acoustic systems. Field-testing of these acoustic recordings systems showed them to be highly functional, with some suggested modifications. The suggested hardware inventory and design for future ARS packages are given in the Appendix.

Appendix 1 ARS Suggested Design

The Acoustic Recording System (ARS) consists of a portable rack-mounted case containing a power supply, amplifier, GPS, computer, and mid- and high-frequency audio interfaces (Table 1, Figure 1). Ships’ AC power is provided to the Astron power supply, which then provides 12v power to the other hardware devices via the fuse box (Figure 2). The Astron must be modified to provide 12v power (Figure 3). Power cables to all units are modified to allow direct connection to a fuse box (Figure 4, 5). The amperage of the fuse will vary per device (see device manuals for specifications). We also suggest minor modifications to the NI-DAQ to decrease noise found to ‘leak’ between terminals (Figure 6). A support was built to provide support to hold the computer, NI-DAQ, GPS, and fuse box; we mounted our support on a tray to provide easy access

(Figure 7). The tray had a simple locking mechanism (Figure 7).

Table 1. Suggested hardware components (make/model) to be used in acoustic recording systems (ARS) for shipboard surveys using towed hydrophones for passive acoustic monitoring of cetaceans.

Figure 1. Suggested ARS Setup including: Astron power supply, Magrec amplifier, Fireface and

NI-DAQ audio interface, GPS, and Logisys computer.

Figure 2. Anterior view of ARS showing lower level sliding tray with fuse box. The orange cable leads the 12v output power from the Astron to the fuse box, which supplies 12v power to each device along with an in-line fuse.

Figure 3. Modification of Astron power supply to provide 12 v power. Unplug device, remove cover, and connect voltmeter between terminals. Using a screwdriver, turn the white knob all the way counter-clockwise, then slowly turn knob clockwise until you reach 12v. It is important to not overshoot, as it takes a little time for the voltage to stabilize. If you overshoot, you may need to turn off the unit and begin again.

Figure 4. Modification of power cables for hardware devices. Each device should come with a power cable (device plug � AC power adaptor � AC plug). The ¼” insulated paddle connectors should be crimped/spliced in-line between the device plug and the AC adaptor. This allows the unit to be connected to 12v power via the fuse box (using the device plug � paddle), or to use AC power (device plug � paddles � AC power adaptor � AC plug). For consistency, always use the male (paddle) connector for the positive (+) supply voltage and the female (socket) connector for the negative (–) supply and label the connectors (see Fig. 5).

Figure 5. 12v power can be supplied to the devices by plugging the ‘+’ and ‘-‘ paddles into the fuse box.

Figure 6. Suggested modifications to NI-DAQ to minimize noise. Small amounts of noise has been found to leak between terminals. We found that separating channel inputs by an unused terminal decreased noise, as well as capping unused terminals with end caps.

Figure 7. Tray modification to provide platform and easy access to computer, NI-DAQ, GPS, and fuse box.