Qualification of DFAN Field & Acoustic Control Methods
Dr.-Ing. Thomas Hoffmann
Acoustic Research Systems
There exist several approaches to control and qualification of Direct Field Acoustic Noise (DFAN) environments, with various degrees of stability, safety, and convenience. This paper aims to compare the methods available today and to expound upon previous claims from the authors. Additionally, the next generation of DFAN testing is further examined to include an expanded consistency analysis as well as a measurement microphone study to benefit quality-assurance of the acoustic field for any test article size.
When performing direct field acoustic testing with electrodynamic devices, methods for identifying, and eventually defeating standing waves in the acoustic field are frequent topics of research. A typical reverberant chamber environment is incoherent, diffuse, and uniform across most of the audible frequency spectrum (1; 2; 3). Since the inception of direct field acoustic testing with loudspeakers, the quest to create an acoustic field that meets or even exceeds the uniformity, incoherence, and diffusivity characteristics of typical reverberant environments has been undertaken by many (4).
Standing waves can create narrow spikes or dips in an acoustic field at various frequencies, leading to excitation of the device under test that is uneven, unrealistic, and potentially even destructive to the hardware being tested (5). Standing waves can typically be seen in sound fields where high coherence exists between two acoustic sources, so historically the narrowband MIMO controllers typically used for direct field acoustic testing have focused on manipulating the coherence of a generated sound field between these points to try and reduce the coherence of the sources and mitigate the creation of standing waves in the field. (6) While this approach has shown some ability to mitigate standing waves and undesirable interferences in the field, there is limited published data on the behavior and characteristics of the acoustic field outside of the control points.
Direct field acoustic testing has been typically performed with repurposed concert sound systems, including vertical line arrays and subwoofers deployed into circular, symmetrical configurations around hardware under test. (7) These devices offer a good balance of output power vs. weight, as well as the ability to be flexibly transported and deployed at various locations. However, the acoustic devices used in concert line arrays must be fully acoustically coupled to neighboring devices to function, and so they require high levels of coherence between sources to generate an adequately powerful acoustic field for direct field acoustic testing.
To create a direct field acoustic testing system that would generate an acoustic field that is as incoherent, diffuse, and uniform as possible, a sound system requiring high coherence would not be the ideal acoustic source to meet those characteristics. To better meet these desired characteristics, a new, purpose-built device was developed, which provides the ability to feed each full-range acoustic source with a drive signal that is completely randomized and uncorrelated from neighboring acoustic devices. This allows for significantly reduced coherence between acoustic devices in both the vertical and horizontal axes, as well as the ability to eliminate standing waves and undesirable interferences from the acoustic field.
To further validate not only the removal of standing waves in the field by a system of purpose-built acoustic devices, but also further explore the characteristics of the acoustic field generated by this system, two significant studies were undertaken.
For these studies, the acoustic testing system utilized was comprised of purpose-built, full-range devices that can be used to emulate legacy acoustic testing methods, as well as generate acoustic fields using a new approach. Following detailed reviews of the topics of vertical coherence and field uniformity in direct field acoustic testing systems, data will be presented to show the results of utilizing 1/3 Octave band control with uncorrelated drives to create an acoustic field with purpose-built acoustic devices, and suggestions for industry-wide best practices will be shared based on the results of these studies.
Vertical Coherence Study
To reduce standing waves in acoustic fields, traditional narrowband control methods for previous-generation direct field acoustic testing systems sought to feed discrete stacks of vertically-arrayed loudspeakers with various mixes of different semi-coherent drive signals. This enabled operators to take advantage of the increase in maximum OASPL that coherence provides within in each stack, a requirement considering the need for an isophasic wavefront for proper line source behavior, while enjoying a measurably diffuse field horizontally, left to right as measured between horizontally arrayed microphone pairs. Mixing various drive signals in this way is effective when measurement microphones are arranged horizontally, but the nature of concert line array type systems requires vertical coherence to produce enough power to deliver launch-level acoustic environments, so it was posited that incoherence in the horizontal plane was adequate to reproduce a chamber-like environment, while maintaining full coherence for each stack in the vertical plane. This of course contradicts the nature of a true reverberant sound field where sound intensity from all directions is uniform, not only in the horizontal plane.
Since one benefit of the RFAN environment is that it is diffuse and incoherent in both horizontal and vertical axis, it was proposed that a study be undertaken to examine how introducing “three-dimensional” incoherence to a direct field acoustic test would affect the acoustic field. To study the effects of vertical decoupling, a series of five acoustic devices were stacked in an arrangement that allowed two control methods: Coherent and Incoherent. The coherent drive method replicated the way that a single concert line array stack would be driven with line-source based systems today, with a single drive signal being duplicated and fed to each acoustic device in the vertical column, and no modifications made to the drive signal phase, equalization, timing, etc. The incoherent drive utilized multiple unique and completely uncorellated noise source outputs, with the number of outputs matching the number of acoustic devices in the column, and each unique drive fed discretely to a single acoustic device. In this way it could be ensured that there was no dependence between each acoustic device in the column, from an audio standpoint, and the control algorithm was optimized to control each acoustic device in the column separately, driving towards an overall reference target for the entire column.
A series of microphones were placed as shown in Figure roughly thirty-two inches from the face of the device. Microphones A1, A3, A5, A7, and A9 served as the control microphones, each one roughly on-axis with the horizontal and vertical center of each acoustic device while this centered placement is not required it was the simplest way to ensure consistency between test runs). Microphones A2, A4, A6, A8, and A10 were placed between the acoustic devices as part of a separate study to characterize the behavior of the column at the relative extremes of each acoustic device, and to demonstrate that the acoustic field could be stretched evenly across the full height of the column without creating hotspots in the middle of the column or inconsistencies in coverage at the edges.
Figure 1 – Vertical Coherence Test – Microphone Placement
The first part of the study was undertaken with the coherent drive approach, the entire column being fed by one single drive signal, as typical of concert line-array systems are driven today. The resulting measurements, Figure 2, showed a number of features that are typical of vertically coherent systems today. First, a ~3dB spread in measured 1/3oct energy below 125Hz, indicating the buildup of strongly coherent low frequency energy interacting with the floor and being measured by the lower microphones, while a notable absence of low frequency energy was measured in the microphones at the top of the vertical array indicating a lack of low frequency buildup and minimal floor-based reflections. Second, fairly good uniformity is seen above 125Hz when viewed in the 1/30ct plots, but that homogeneity is not necessarily maintained as well when the same data is viewed in 1Hz resolution, Figure 3, indicating some potential issues with vertical standing waves or “hotspots” in the vertical acoustic field (e.g. the dip at 350 Hz or the equally spaced peaks above 1000 Hz). At these frequencies the controller can’t reach the specified amplitude due to nodes of the room modes at the control microphone locations.
Figure 2 shows a 60 second linear average of sound field exposure at the five control microphone locations, in 1/3 Octave resolution. Of note are the dips in the frequency response of microphone seven at the 1600Hz and 2000Hz 1/30ct band, the result of destructive interference, theorized as standing waves in the vertically coherent array. Additionally, the uniformity of the measurements below 125Hz is noteworthy, as it shows a significant deviation in LF uniformity from the microphone at the bottom of the array to the microphone at the top of the array.
Figure 2 – Vertically Coherent 1×5 Stack – Control Mics. Octave
Figure 3 is the same information as Figure 2, but processed at 1Hz resolution to show a higher degree of detail in the microphone frequency responses. The same destructive interferences around 1600Hz and 2000Hz are more clearly visible, as well as narrower non-uniform responses in the microphone responses that might have not have been clearly seen in the broader 1/30ct resolution plot.
Figure 3 – Vertically Coherent 1×5 Stack – Control Mics. PSD
The second part of the study was undertaken with the incoherent drive approach, each acoustic device in the column being driven discretely with its own uncorrelated random noise source output signal, and that drive being controlled by an individual microphone placed in front of the acoustic device in its direct field. When compared to the previous results of the coherently driven study, addressing the same two key points noted, the incoherent drive shows significant improvement to both concerns. In Figure 4 there is no build-up of low frequency energy at the bottom of the vertical column, indicating that the effects of the ground reflection or “floor bounce” have been mitigated by the individual, uncorrelated drives for each acoustic device driving to the reference target and utilizing the 1 second control loop averaging time to account for the acoustic environment seen in nearly real-time. In Figure 2, where all devices were driven correlated, controlling for the floor reflections would have negative impact on the sound field further up the stack. Second, when viewed as a narrowband spectrum, Figure 5, the improvements to uniformity are also seen, as vertical standing waves have been removed from the column and the vertical acoustic field is more diffuse and uniform. These acoustic field characteristics bear resemblance to previous published work by the authors where the incoherent drive method was utilized with a horizontally deployed (circular) system of acoustic devices. (8)
Figure 4 shows a 60 second linear average of sound field exposure at the five control microphone locations, in 1/3 Octave resolution using the incoherent drive method. Compared to the same plot from the coherent drive method, overall responses are more uniform broadband. There is an increase in LF energy in the lowest microphone, seen below 100Hz in the A1 microphone response, as well as increased non-uniformity in microphone A7 vs. the rest of the control microphones.
Figure 4 – Vertically Incoherent 1×5 Stack – Control Mics. Octaves
Figure 5 is the same information as Figure 4 but processed at 1Hz resolution to show a higher degree of detail in the microphone frequency responses. Of note with this incoherent drive method is the lack of narrow destructive interferences in the high-resolution data that was seen in the control microphones capturing the coherent drive method.
Figure 5 – Vertically Incoherent 1×5 Stack – Control PSD
Figure 6 shows that by averaging the narrowband spectra of the control microphones the presence of vertical standing waves can be hidden. The blue spectrum shows the average of all control microphone spectra from Figure 3, whereas the red line similarly shows the average of Figure 5. Despite major differences existing in both acoustic fields regarding diffusivity, as shown above, just by looking at the averages it appears as if there was no significant difference between both methods. It is therefore strongly suggested by the authors to judge the diffusivity of the sound field by means of individual microphone spectra.
Figure 6 – Average Control Mics Comparison Plot
Figure 7 – 1×5 Coherent Stack – Coherence Plot
Figure 8 – 1×5 Incoherent Stack – Coherence Plot
Analyzing the coherence plots for both drive methods, fully correlated and uncorrelated, Figure 7 and Figure 8 clearly illustrate the difference between both. Shown is the coherence function between microphone 1 and every other microphone, see Figure 1. The coherence of microphone pairs allows judging the diffusivity of the acoustic field.
For correlated drive signals the acoustic device column acts as a line source with a uniform wavefront resulting in high coherence. For higher frequencies and smaller wavelengths mixing of wavefronts and effects like reflection and scattering are more prominent resulting in a “noisy” coherence spectrum with occasional dips. Another effect is the excitation of room modes, where at the coherence dips a node of a room mode lies. In nodes, the coherence becomes naturally low as
Here, C is the coherence and S is the auto- and crosspower spectrum of microphones x and y. If one of those lies exactly in a node of a room mode, the crosspower Sxy becomes zero and therefore the coherence experiences a dip. The overall high coherence still gives away the correlated drive method.
For uncorrelated drive, Figure 8, the overall coherence between microphone pairs is low as expected. For low frequencies (below 50 Hz) for pair A01 and A03 the relatively high coherence results from the fact that the wavelength (150Hz = 5.6 ft) is much bigger than the distance between microphone pairs, which makes it physically impossible to create fully uncorrelated wavefronts. At higher frequencies the inherent randomness of the field causes peaks in the coherence spectrum at certain frequencies. Still, the field can be described as fully uncorrelated as the overall trend of the coherence is more significant for a random signal than possible sharp peaks at few frequencies.
Comparing the average coherences for all microphone pairs, Figure 9, clearly shows the different nature of both drive methods and further underlining the previous discussion.
Figure 9 – Coherent vs Incoherent 1×5 Stack – Averaged
Standing waves in the sound field can also be made visible by analyzing the transfer function between two microphones, that is
Where S is the PSD of microphones x and y. Figure 10 and Figure 11 show those for microphones 3 and 7 for uncorrelated and correlated drive respectively. The resonance peaks visible in the transfer function with correlated drive signals give away hotspots in the acoustic field. These can be interpreted as acoustic modes between the spatial locations of microphone pairs. In this example, there is a very prominent mode at 1.5 kHz which might result in an acoustic hotspot. Placing a DUT in that hotspot could lead to overtesting and should be avoided. Using uncorrelated drives there are no resonances visible, further proving the benefit of this method.
To fully analyze the acoustic field for correlated drive and making sure no hotspots are in critical locations (e.g. flexible solar arrays of antennas), an infinite number of microphone pairings randomly distributed in the field would be necessary to find all local modes. As this is unfeasible, also from a controller algorithm speed standpoint, having uncorrelated drives for each acoustic device is the preferred method.
Figure 10 – Control 3 vs. Measure 7 – Incoherent
Figure 11 – Control 3 vs. Measure 7 – Coherent
Figure 12 – Superimposed Transfer Function derived standing wave with Coherence Plot
From the authors standpoint, there are two key points to be made from the results of this vertical drive methods study, which are highly relevant to conversations about direct field acoustic testing methods today.
First, the identification of standing waves in the vertical axis seems to indicate that the effects should be quantified with regards to possible effects on structural excitation. Over the history and evolution of direct field testing methods and technologies, a great deal of effort has been undertaken to ensure that standing waves were mitigated or eliminated in the horizontal plane, as the presence of standing waves in the field using early SISO methods – as an example – was typically described by authors of early papers on the method as a serious roadblock to the progression and acceptance of direct field acoustic testing as a reliable and accurate testing method. (9) Based on that legacy, the identification of these same standing waves but in the vertical axis seems to indicate more work needs to be done to examine the efficacy of vertically coherent systems.
Second, in line with previously published studies (8) on utilizing incoherent drive methods, coherence measurements of multiple acoustic devices driven by discrete uncorrelated drive outputs show in both axes not only significantly improved (from a structural excitation standpoint), consistent broadband incoherence as seen in reverberant chambers, but also low frequency coherence behavior (below 100Hz, typically) that improves on the typical acoustic field found in both reverberant chamber and modern direct field environments. This is a point that must be studied further, as generation of a “reverberant chamber-like” acoustic fields were always seen as a minimum-viable target for direct field acoustic testing systems.
From a structural excitation standpoint, when prompted with this improved testing capability, some end users of incoherent direct field acoustic testing systems have responded positively to the ability to test with a more broadband, more randomized field as part of their test regimen. This type of field was even preferrable in some cases to classical reverberant chambers, which exhibit the same low-frequency coherence as direct-field systems which employ vertically coherent line arrays.
Though it is known that vehicle fairings do not produce a fully diffuse or evenly distributed field, the capability to create one in a laboratory environment provides a highly repeatable test that can serve as a consistent, homogenized test reference as well as envelope any possible combination of fairing environments. On the other hand, drive methods and control strategies that are not able to create fully diffuse sound fields independently of microphone setup will have a hard time to simulate sound fields as seen by payloads in various launch vehicle types and configurations without advanced beamforming techniques that will serve to further reduce output levels. Furthermore, it would be an incredibly complex and impractical undertaking to attempt to capture and emulate the nuanced differences in partial field diffusivity between all the different fairings and environments in the market today and in the future, with little added benefit to the test itself.
Expanded Consistency Study
A previous consistency study (8) aimed at characterizing the uniformity of the field for a system that utilized fully incoherent sources for each acoustic device. When presented with the argument that direct field acoustic testing systems must be prescribed to certain size diameters and distances from microphones to enable consistent phase-alignment, the authors designed a study that showed through fixing measurement microphones around a mock test article, the diameter of an incoherently-driven system that does not control using phase could be modified almost at will without having any effect on the quality or uniformity of the acoustic field at the fixed microphones (and DUT).
Via peer review, questions were raised regarding the quantity of microphones utilized in the study, and whether that quantity was adequate when compared to current industry guidance. To further build on the existing dataset, as well contribute to the growing discussions on microphone quantities in acoustic testing applications, the authors undertook a more in-depth consistency study to explore these topics further.
The two main points to be explored go together, namely, can the field be studied in greater detail with more datapoints? And, if the field is being driven properly with incoherent sources, does having too few or too many microphones affect the quality of the test? Regarding the latter question, historical guidance on the topic has been that 16 was the correct number of control microphones to use for a test (HDBK-7010), and that was later revised to 24 being the correct number of control microphones (citation) to use for a direct field acoustic test.
To further study this topic with expanded horizons and provide expanded guidance for direct field acoustic control systems that not only did not utilize MIMO-narrowband acoustic control methods, but also could incorporate large numbers of non-control monitoring microphones for field QA purposes, this expanded consistency study was undertaken.
First, a direct field acoustic test system was installed around a mock test article. Like the previous study, the acoustic devices would be relocated to progressively more distant positions to create larger diameter test circles, while the monitor microphones (not in the control loop) meant to characterize the field would remain in fixed positions to measure the field after each configuration change. The author’s target for this study was to provide 92 microphone inputs for each diameter of the test configuration, both to increase the detail vs the previous study by a significant factor, and to alleviate any potential concerns from the scientific community that too few microphones were used to control the system and study the field.
The microphone configuration consisted of both control and monitoring mics. Each acoustic device utilized a single control mic placed roughly half the distance between it and the test article, for a total of twelve. Two sets of eight monitoring microphones were attached at a vertical beam spanning from bottom to the top of the acoustic field with slightly randomized locations. Those beams can be moved freely
The steps for the study were as follows:
1. The setup was configured for diameter #1
2. Microphones were placed
A. Control microphones placed at each acoustic device
B. Roving monitor microphones placed at position A
C. Fixed monitor placed at their fixed positions
3. The 1540E acoustic profile was entered and executed at full test level for 60 seconds
4. Roving monitor microphones were moved from position A to position B
5. Repeat steps 3 & 4 until roving monitor microphones had been placed at positions A-E
6. Reconfigure setup for next larger diameter
7. Repeat steps 2-4 until all positions have been captured
8. Repeat entire process for each diameter
Following these steps allowed the authors to conduct the same exact acoustic test multiple times, and for each test move the roving monitor microphones to 5 predetermined locations. Therefore, for each diameter of test configuration, with 36 inputs available, 92 microphone positions could be captured. In Figure 13, below, the (12) control microphones (green dots) and (80) monitor microphones can be seen together, illustrating the ‘cloud’ of field verification.
Figure 13 – Consistency Study – All Microphone Locations
Overall, the results of this study reveal similarities in diffusivity of field, but with somewhat larger variation in uniformity for microphones not in the control loop.
The plots below are a linear average in both 1/3 octave and 1hz PSD for 60 seconds of exposure for all twelve control microphones across all five individual runs, Figure 14. These plots assure that from one run to the next, no significant variation occurred which would affect the roving placement of monitor microphones. It’s worth noting that no additional spectral averaging or smoothing were applied to any of the plots in this section, unless otherwise specified.
Figure 14 – 12′ Diameter – Control Microphones
As noted in previous studies and regardless of the type of control system used, control microphone locations will always follow the reference spectrum more closely, as they are actively used to shape the field. Looking outward in all directions is more revealing and can help to qualify the true diffusivity of the sound field. It was expected that these monitor locations exhibit a wider spread with respect to amplitude but follow the same basic shape (uniformity) and relative diffusivity.
The plots below show the (80) individual non-repeating monitor locations and are believed to map the quality of field more accurately, Figure 15. As previously noted, the control system’s tendency to maintain energy generally at or below the reference spectrum had a direct impact on the monitor locations’ measurements. The average of the measurements trended about 2dB below nominal. Though this is within typical measurement acceptability, it’s worth noting that a high percentage of the monitor microphones were placed further away from the acoustic device than the control microphones and were therefore expected to trend lower.
Figure 15 – 12′ Diameter – Monitor Microphones
Seeking to challenge previously published methods for displaying datasets from direct field acoustic test field qualification studies, Figure 16 shows the average of the control microphones for one run plotted alongside the average of the monitor microphones for the same run. Speaking visually, this presentation is very appealing, as it shows both the control and monitor microphones in good agreement with the target reference trace. However, these charts also both represent the combined averaging of nearly 92 microphones (12 control, 80 measurement) into a pair of lines. With this in mind, it is difficult to see the benefits of highly averaged datasets for presentation when the measurements shown are meant to clearly elucidate the diffuse and uniform nature of a large acoustic field at a diverse number of independent points.
Figure 16 – 12′ Diameter Avg. Control vs. Monitor
To verify that the field measurements are not relative to a particular test article size or test diameter, the data presented below are from a test where all things were identical except the diameter of the test circle (15′ instead of 12′) and the control microphone locations. The latter were moved to maintain a randomized location pattern (+/-8″) which was roughly half the distance between the acoustic device face and the test article but varied in height. It was observed that the data obtained from the second configuration was generally identical to that of the previous, with a similar trend of roughly -2dB, Figure 17 – Figure 19
Figure 17 – 15′ Diameter – Control Microphones
Figure 18 – 15′ Diameter – Monitor Microphones
Figure 19 – 15′ Diameter Avg. Control vs. Monitor
The data from this study helped to understand that adding control microphones to average the individual loops more generally serves to draw the control and measurement fields more closely together. Subsequent process improvements made for industry tests on flight hardware have shown that operating in this way with fully incoherent sources provides a quality-of-field and excitation similar to or better than that of a reverberant chamber but using significantly less resources.
Having answered the first question posed in this improved consistency study, can the field be studied in greater detail with more datapoints, the second question is to be addressed. Namely, if the field is being driven properly with incoherent sources, does having too few or too many microphones affect the quality of the test? An answer is proposed by the way of offering suggestions for best practices when approaching a next-gen type of direct field acoustic test.
Based on the data seen here and in previous studies, the authors put forth the suggestion that rather than declaring one quantity of inputs to be the correct number, direct field acoustic testing vendors and end users seek to understand the control method of the acoustic test, and deploy:
- A quantity of control microphones that provides adequate capability to properly control the test per the vendor’s control strategy and establish that the acoustic field is being managed both horizontally and vertically around the entire volume dedicated to the test article.
- A quantity of non-control monitor microphones in various positions around the acoustic field that act as a quality assurance method for both vendor and end-user.
These non-control monitoring microphones provide not only real-time OASPL and frequency response data for display during testing, but also accurate and aligned (to test schedule) time histories for post-processing purposes, enabling strong confidence that the field is correct not only at the control microphone positions, but at many positions around the volume dedicated to the test article as well.
Further to the point of providing confidence in measured acoustic data, the methods undertaken in this paper have been reviewed carefully and compared to reverberant environments by end users with both simulators and flight hardware. In all cases, an appropriate number of control and monitor microphones scaled for UUT and field size along with simple post-processing methods applied without complex
averaging or smoothing algorithms, revealed adequate excitation. While it is important to employ monitor microphones for quality assurance purposes, it is arguably equally important to post-process and review the dataset provided by those microphones with simple linear average methods that show not only the entire exposure of the test article to the full test level, but also review various 5-10 second snapshots of linearly averaged timeframes from beginning to end to ensure the test ran stably and responsively
In the ongoing quest to make direct field acoustic testing a more accessible, standardized, and repeatable process, the creation of a fully uniform and diffuse sound field continues to be a worthwhile objective. New, purpose-built direct field acoustic testing systems that enable the use of multiple, uncorrelated drive signals have shown promise in delivering a sound field that is as well suited for acoustic spacecraft testing as those found in reverberant chambers.
It was shown that for vertical columns of acoustic devices, uncorrelated drives for each acoustic device are the preferred method to avoid safety-critical, localized hotspots. Conventional line-array type DFAN testing systems which utilize many loudspeakers apparently suffer from this effect, as they lack efficiency and power to operate independently, and both low- and high-frequency lobing can be seen.
In the review of field uniformity, it was shown that by qualifying the field with an many randomly placed monitor microphones, a great deal of additional information can be captured which can be helpful to assume similarly designed configurations will produce repeatable results. Additionally, while helpful for revealing generalized trends, it was seen that capturing and displaying the results of many measurements in an average may obfuscate field quality and uniformity. It is therefore important to avoid smoothing or averaging of responses when monitoring a field.
The results of these studies seem to warrant further study into the performance of fully incoherent direct field acoustic testing systems with purpose-built acoustic devices, with regards to uniformity and incoherence, in larger deployments and with an increased emphasis on the analysis of structural excitation within the test circle.
- Design and Instrumentation of a Large Reverberation. Peterson, Jeffrey Shawn and Bartholomae, Roy C. Cleveland, Ohio: s.n., 2003. Proceedings of the 2003 National Conference on Noise Control Engineering.
- Nolan, Melanie, Berzborn, Marco and Fernandez-Grande, Efren. Isotropy in decaying reverberant sound fields. Journal of the Acoustical Society of America. 148, 2020, Vol. 2.
- Grosveld, Ferdinand W. Characterization of the Reverberation Chamber at the NASA Langley Structural Acoustics Loads and Transmission (SALT) Facility. s.l. : NASA/CR-2013-217968, 2013.
- Hargreaves, Jonathan. Literature review of Direct Field Acoustic Noise (DFAN) testing. Salford : University of Salford, 2022.
- Kolaini, Ali R., Doty, Benjamin and Chang, Zensheu. Impact of Acoustic Standing Waves on Structural Responses: Reverberant Acoustic. Pasadena: Jet Propulsion Laboratory, California Institute of Technology, 2012.
- Comparisons of the Structural Response of a Test Article Excited by DFAT(TM) Diffuse and Non-Diffuse Acoustic Fields. Marcos A. Underwood, PhD. Orlando: s.n., 2021. 91st Shock and Vibration Symposium.
- NASA TECHNICAL HANDBOOK: NASA-HDBK-7010. s.l. : National Aeronautics and Space Administration, 2016.
- Verifying a Novel Approach to Acoustic Testing: Acoustic Devices & Control Systems. Shick, et al. Virtual: s.n., 2021. 32nd Aerospace Testing Seminar.
- Stasiunas, et al. Experiences in Performing a High Intensity, Direct-Field Acoustic Test on a Contamination Sensitive System. Albuquerque: Sandia National Laboratories, 2012.
About the Authors
Dale Schick is Applications Manager of m+p International Inc. and leads US research projects related to Acoustic and Vibration testing. He has extensive experience in environmental testing as applied to Aerospace and over 15 years of practical experience in the industry.
Thomas Hoffmann is product manager of m+p international and coordinator of the Direct Field Acoustic Noise Testing solution. Before he joined m+p international he worked as a researcher at Institute of Dynamics and Vibration Research in Hannover, Germany. There he received his PhD degree on experimental nonlinear vibration and simulation algorithms.
Eric Friedlander is the CTO and Co-Founder of Acoustic Research Systems. He has a BS in Audio Technology from American University in Washington, DC, and over two decades of practical experience in sound engineering for musical, industrial, and scientific applications. Prior to founding ARS, Eric was a product manager at Harman Professional for the JBL Professional Tour Sound portfolio, and prior to that he worked as an acoustic testing engineer at MSI-DFAT.