
A measurement device for sound field sampling in rooms
Master Thesis of Thevißen, Florian
Forschungsgebiet: Raum und Bauakustik
Betreuer: Witew, Ingo / Vorländer, Michael

A waveletsynthesis model for the auralization of moving sound sources
Master Thesis of Wenhuan Duan
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Meng, Fanyu
Moving sound sources are indispensable parts in acoustical environment and cause noise problems as the speed of the sources increases rapidly, such as highspeed trains and cars. Auralization is an efficient way for prediction and provides immerse audible sense. Therefore, it is important to find a sound synthesis approach to modeling moving sound sources and create perceptually convincing sounds for auralization. In this research, wavelet transform (WT) is proposed to synthesize moving sound sources in virtual reality environment. Compared to Fourier transform, WT has the advantage of remaining time information when transforming signals into the frequency domain, which benefits the received timevariant signals from moving sources. Due to frequency shift, the Doppler effect should be eliminated through interpolation before WT. With varying the parameters generated from WT, new signals can be synthesized based on several samples. The results need to be verified by onsite measurements listening tests. Finally, a synthesis model for auralization is established based on WT for moving sound sources.

Acoustic Surface Impedance Estimation with a Hybrid Measurement and WaveBased Simulation Method
Master Thesis of MüllerGiebeler, Mark
Betreuer: Vorländer, Michael / Opdam, Rob
Existing and commonly used methods to measure the sound reflection properties of acoustic surfaces have certain restrictions. Either they inherently do not provide enough information (for example phase information for wave theory based simulations), or they only yield realistic results to a limited extent (for example in a perfectly diffuse sound field, under a plane wave incidence assumption or only for a limited frequency range) . More sophisticated methods to determine the complex angledependent reflection factors are often complicated and very timeconsuming.
This work presents an inverse method that only needs a single sound pressure measurement of a finite porous absorber sample, along with geometric information for simulation, as input data and takes into account the actual incident sound field as well as a potentially nonlocally reacting material. With a nonlinear fitting algorithm, the simulated complex pressure data is adjusted to match the measured data by changing the absorber model parameters (flow resistivity, porosity, etc.). Several factors that affect the measurement and/or the optimization process are investigated theoretically. Furthermore, an extension to the above mentioned approach is proposed, allowing for a edge effect correction of the finite material probe. Using iteratively refined FEM simulations that are based on the same geometric dimensions of the absorber sample as in the measurement, enables to compensate the introduced error and to determine the impedance as measured on an infinitely extended material probe. The method is validated based on simulations and applied in preliminary measurements.

Adaption of room acoustic auralizations to the reproduction environment
Master Thesis of Voth, Markus
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Kohnen, Michael
This master thesis investigates how far virtual rooms can be adapted to reproduction environments. To achieve this, the properties of early and late reflections are used with respect to physics and hearing physiology.
In case of insufficient absorptive materials, early reflections and long reverberation times occur in rooms. Thus, the reproduction of the virtual scene is distorted.
To compensate for reverberation, there are various approaches. For early reflections, it has been shown that a wave based annihilation is highly sensitive to small disturbances. Therefore, in this thesis, early reflections of the reproduction room shall be integrated into the VRscene as good as possible by using a mapping method. To visualize early reflections, a vector based figure is developed.
Objective evaluation criterions or measures of error are explored to compare room impulse responses. To reduce these errors, the above described compensation strategies are developed. Though, compensation possibilities are limited by the room where signals are reproduced. To evaluate the results objectively, there are simulations and measurements. For the subjective evaluation, a headphone based listening test is executed.

Analysis and Design of a Matched Microphone Array for MIMO Applications in Room Acoustics
Master Thesis of Berzborn, Marco
Betreuer: Vorländer, Michael / Klein, Johannes
Arrays of microphones and loudspeakers can be utilized to study spatial properties of acoustic
wave fields in rOoms. For example, microphone arrays can be utilized to estimate the direction of
sound incidents corresponding to single reflections, while loudspeaker arrays are capable of exciting
single reflections in a room. Therefore, the combination of both to an acoustic mUltipleinput
multipleoutput (MIMO) system provides a powerful instrument for room acoustic analysis.
The first part of this thesis focuses on an analytic formulation of such a MIMO system. In particular
potential error sources such as spatial aliasing and model mismatch and their impact on the
formulated system are studied. Based on the formulation of errors, a simulation framework is
derived, aiding at the determination of a MIMO system that inherits the lowest total error for a
given source array. F inally, a system fulfilling the minimal error criterion is identified. This thesis
then concludes with the design of the spherical microphone array based on the results of the error
simulations.

Analysis of Directional Decay Curves Based on Simulated Room Impulse Responses
Bachelor Thesis of Bilitewski, Niclas
Betreuer: Berzborn, Marco / Klein, Johannes

Analysis of Room Impulse Responses Measured with Compact Spherical Microphone and Loudspeaker Arrays
Master Thesis of Haar gen. Epping, Christian

Atmospheric Ray Tracing based on altitudedependent weather data
Master Thesis of Schäfer, Philipp
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Mecking, Jens / Stienen, Jonas
This thesis addresses sound propagation in the atmosphere in the context of auralization of acoustic scenes. A common method to estimate sound paths between a given source and receiver is the so called ray tracing. With the help of these estimated paths, an impulse response can be derived, which allows the auralization of acoustic scenes in a virtual environment. In contrast to the acoustic medium in rooms, where ray tracing is commonly used, the atmosphere is neither homogeneous nor static. Main reasons for this are the variation of the sonic speed over space and movement of the medium, such as wind. These effects lead to refraction and translation of sound resulting in curved ray paths. Additionally, distances between source and receiver are rather large in outdoor scenarios. Therefore, ray density around the receiver is low and prediction of ray paths needs high computational effort.
This impedes finding eigenrays  rays, that directly hit the receiver  since it is complex but needs to be efficient for fast computations of the impulse response.
In the course of this thesis, a ray tracing algorithm for the simulation of atmospheric sound propagation and auralization is designed. Ray propagation is investigated to determine criteria for finding and neglecting rays, that are irrelevant for the receiver, in an early state to save computation time. Altitudedependent weather data is used to model atmospheric properties assuming the atmosphere to be a stratified medium. In a second step, the influence of these properties on the impulse response is investigated. Therefore, an acoustic scene is rendered repeatedly while varying the weather parameters.

Auralization of Building Acoustic Filters
Forschungsgebiet: Raum und Bauakustik
Betreuer: Imran, Muhammad
In building acoustics auralization, the implementation of sound transmission (i.e. airborne and structure borne transmission etc.) is an important phenomena to be incorporated in virtual buildings from where we can evaluate the performance of the building elements in terms of noise and comfort. The sound transmission through building elements influence the perception to a listener in listening rooms, therefore, there is a requirement to design the filters for building elements as acoustic sources to predict their influence. In the receiving room these building elements act as secondary sound sources and these secondary sources are obtained from transfer functions of sound transmission through building structures. This research topic focuses on handling the virtual building geometries and rendering the secondary sound sources in addition to the room acoustic filters to auralize the perception of the buildings under study. The potential applications of this research are the auralization of building in virtual reality and video games.
Requirements
? Interest in the topic and independent work
? Knowledge of Room Acoustic
? Familiar with Matlab
Optional skills
? C# Skills
? Working in Unity

Auralization of Sound Insulation in Virtual Reality
Forschungsgebiet: Raum und Bauakustik / Akustische Virtuelle Realität
Betreuer: Imran, Muhammad
In this research, we investigate the auralization of airborne sound transmission in complex buildings to develop the corresponding auralization filters chain for the evaluation of the performance of these building in terms of noise and comfort. This study focuses on the implementation of airborne sound transmission based on ISOEN: 123541, and comprehends the calculation procedures for sound insulation metrics (i.e. sound reduction index for direct transmission of the different structures and the vibration level differences across junctions) and the development of sound insulation filters. These filters calculates the sound transmission between dwellings by partitions and by flanking structures to estimate the transfer functions between the sources and receivers during auralization process. Example buildings would be taken as a test case that consists of different type of building elements and their constructions. These buildings would be presented in virtual reality and the insulation filters will be applied to different scenarios (i.e. different source and receiver positions in different coupled rooms) and as a result synthesized room impulse responses (RIRs) will be obtained for these scenarios. Perceptual studies will be conducted to evaluate the performance of these buildings and to predict the comfort and annoyance. The potential applications of this research are the auralization of building in virtual reality and video games.

Automatische Geometrievereinfachung für Raumakustiksimulationen
Bachelor Thesis of Durand, Christopher
Forschungsgebiet: Akustische Virtuelle Realität / Raum und Bauakustik
Betreuer: Aspöck, Lukas

Building acoustics simulation based on semantic models
Forschungsgebiet: Akustische Virtuelle Realität / Raum und Bauakustik
Betreuer: Stienen, Jonas / Imran, Muhammad
Room acoustics simulations are mostly based on geometry meshes with polygonal faces that are linked to acoustic materials, such as absorption and scattering coefficients. Unfortunately, building acoustics simulations require also sound transmission through walls and ducts. Here, polygonal meshes are inherently insufficient, because solid structures can not be described efficiently. In building and architectural modeling, semantic data structures based on the BIM approach are widely used and well integrated into the design and model workflow.
In this master thesis, the employment of IFCcompatible semantic input data for building acoustics simulation will be investigated. Ways to extract room meshes for room acoustic simulation are investigated. A derivation of sound transmission paths between rooms will be formulated that can be readily used for building acoustics simulations with flanking paths.
Requirements
? Basic C++ programming skills
? Basic knowledge of technical acoustics
? Basic knowledge of acoustic simulation
? Interest in the topic and independent work
Optional skills
? Familiar with SketchUp, FreeCAD, Revit
? Knowledge of semantic data structures (namely IFC)
? Familiar with Visual Studio

Characterization of simulated jet noise by means of spherical harmonics decomposition
Master Thesis of Cappellotto, Francesco
Forschungsgebiet: Lärmforschung / Akustische Virtuelle Realität
Betreuer: Mecking, Jens
This thesis presents a research on the source modeling of aircraft’s jet noise, the goal of which is to reproduce the sound of aircraft in simulated Virtual Reality environments with a high degree of fidelity.
At the moment, the acoustical modeling of jet noise follows semiempirical models, partially derived from flyover measurements and optimized to include a set of variable configurations, e.g. noise reducing chevrons. However, this method has some limitations. In fact, its spectral resolution is limited to onethird octave bands, and the correlation of single noise components is not defined. Moreover, ground measurements require some approximations to account for the cutback phase (a reduction in the power level right after takeoff). This affects not only the power level, but also the directivity of the jet noise, which cannot be approximated due to lack of information. Finally, ground measurements are normalized to standard atmosphere, thus the uncertainty derived from the current weather at the time of the measurements is included in the directivity.
In the scope of this thesis a method will be researched, to model the noise of jet engines in a
more precise way, using simulated data to compute physically valid directivity patterns, rather than empirical or semiempirical models, and to achieve results that are independent from meteorological conditions and not restricted to onethird octave bands.
To do so, simulated data on a 3dgrid in the near farfield around the jet will be used as input data. A set of points has to be chosen and modeled using spherical harmonics decomposition. Since it is expected that the model will reach highorder spherical harmonics, an acoustic centering of the source will be evaluated to reduce the computational cost.
With the use of this method, it is expected to achieve a better auralization, which can be used for more realistic VR simulations.

Characterizing and Analyzing Auralizations of Complex Acoustical Scenes
Bachelor Thesis of Reffgen, Matthias
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Aspöck, Lukas / Stienen, Jonas

Comparing inverse absorption coefficient methods to traditional reverberation chamber measurements
Master Thesis of Gomez, Lian
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Aspöck, Lukas
Absorption coefficients as a description of the boundary conditions for room acoustics simulations are usually very challenging to determine. A standardized method estimates the random incidence absorption coefficient of a material sample by placing the material in a reverberation chamber, conducting an acoustical measurement and comparing it to another measurement of the empty chamber. As this method has some drawbacks and limitations, a novel approach uses a room acoustics simulation and matches the input parameters until a match is found for a room acoustics measurement of the identical situation. In this work, both methods should be applied for at least two samples and their results should be analyzed and compared. Especially the problems of nondiffuse sound fields and nonlinear decay curves should be investigated.

Continuous measurement of room impulse responses with
Spherical Loudspeaker Arrays
Bachelor Thesis of Palenda, Pascal
Forschungsgebiet: Akustische Messtechnik
Betreuer: Klein, Johannes
The results of a room impulse response (RIR) measurement are used in acoustic research for auralization, room reflection analysis or sound field analysis. When measuring a single RIR, the measurement system has an intrinsic directivity. Under the assumption that the room complies with the lineartimeinvariantsystem (LTIsystem) prerequisites, superposing single RIRs, taken at specific directions, can synthesize a RIR with arbitrary directivity. The current state of the art is to measure the single RIRs sequentially using a spherical loudspeaker array (SLA) and spherical microphone array (SMA). Nevertheless, hundreds of single RIRs have to be measured for a highresolution RIR with arbitrary directivity. Due to the long measurement time, the room is subject to changes in temperature and humidity. This in turn invalidates the timeinvariance of the assumed LTIsystem.
The current goal is to minimize the time to complete the measurement. One proposed way is a continuous measurement. With this method the SLA is rotated continuously therefore we can simplify the problem to a measurement with an SLA and a single microphone. RIRs for different spatial directions, corresponding to the sequential measuring direction, can then be extracted from the captured signal. This would in turn eliminate the time of mechanical free travel and oscillations caused by adjusting the SLA and thus reducing the measurement duration.
This thesis will examine the viability of continuous measurements for RIR measurements. A comparison methodology will be explored, subsequently the results of the continuous measurements can be compared to the sequential measurements. Furthermore, this thesis explores the limitations of a continuous measurements. Different excitation signals will be implemented, simulated, measured and postprocessed using a singleinput multipleoutput (SIMO) measurement setup with an SLA.

Creation of auditory scenes for multimodal listening experiments
Bachelor Thesis of Geusen, Holger
Forschungsgebiet: Raum und Bauakustik / Akustische Virtuelle Realität
Betreuer: Aspöck, Lukas / Vorländer, Michael
Zur Bewertung und Validierung von raumakustischen Simulationen können diese mit Referenzmessungen verglichen werden. Ein solcher Vergleich soll zukünftig mit Hilfe von audiovisuellen Experimenten durchgeführt werden. Sowohl für akustische Simulation als auch für die visuelle Darstellung einer virtuellen Szene gibt es jedoch eine Vielzahl an Lösungen, was Hard und Software betrifft. Aus diesem Grund sollen in dieser Abschlussarbeit verschiedene Tools und Umgebungen für das akustische und visuelle Rendering recherchiert und basierend auf einem erstellten Kriterienkatalog für die Eignung in den geplanten Experimenten bewertet werden. Dazu werden drei Raumszenen jeweils in mehreren Systemen erzeugt und in Pilotversuchen evaluiert.

Development of a measurement method for the analysis of surface vibrations
Bachelor Thesis of Peckert, Fabian
Betreuer: Mecking, Jens / Klein, Johannes
Mikrospeakers, which are normally designed as dynamic loudspeakers, are very compact and can be
found in portable devices. Therefore, many approximations of conventional loudspeakers are only
partially valid for microspeakers. In particular the assumption of a loudspeaker membrane as an
inphase vibrating piston radiator does not apply for the whole frequency domain, since for certain
frequencies vibratingmodes propagate in form of bending waves on the membrane.
In order to investigate the positiondependent surface velocity of a loudspeaker membrane, this
bachelor thesis presents a method to measure the membrane. Furthermore, with the help of the
developed method the vibration pattern of the membrane can be visualized and analyzed. The
scanning of the membrane is done by a laservibrometer that measures the velocity of the membrane.
The measurement position is shifted by with two step motors which are attached to a measurement
table. The measurements take place all over the membrane with an adjustable resolution of the
measurement grid. A broadband excitation signal which covers the whole human perceptible
frequency range is used. This study is limited to the measurements of microloudspeakers, but
can also be extended to analyze other vibrating surfaces. The analysis can be done for arbitrary
frequencies by visualizing the different measurements with 2D or 3D graphic tools. Moreover
transfer functions or impulse responses for individual measurelnent points on the Illmnbrane can be
illustrated in order to study the formation of modes in the surface.
Additionally a method for the creation of a video was made that shows the vibration characteristics
of the membrane for different frequencies. Hereby the propagation of vibrating modes and the phase
differences between input and output signal can be studied and analyzed which can help to modify
microloudspeakers and its radiation characteristics. With regard to the future, the aim is to use
the developed measurement method and the visualization tools for teaching purposes. Hereby the
methods shall be integrated into the 'Akustische Praktikum'lab course and further research on
microspeakers at the Institute of Technical Acoustics.

Development of Sound Insulation Toolbox (GUI)
Forschungsgebiet: Raum und Bauakustik
Betreuer: Imran, Muhammad
The development sound insulation models for airborne transmission are important for evaluating the performance of building structures in terms of noise evaluation. The sound insulation models describe the performance of the building elements. The basic parameters that are necessary for the development of these models are based on measurements and/or on calculations and material properties of the building elements, according to the ISO standards. Once the sound insulation models are developed these models are implemented in graphic software such as Sketch Up and Unity. This research focuses on developing a graphical user interface (GUI) based toolbox for calculating the sound insulation parameters for virtual building geometries in Matlab software and rendering the corresponding sound transmission filters to auralize the perception of the buildings under study. This toolbox will be used for education and research purpose in the area of building acoustics.
Requirements
? Interest in the topic and independent work
? Knowledge of Room Acoustics
? Matlab
Optional skills
? Knowledge of Building Acoustics

Diffraction simulation in Geometrical Acoustics
Master Thesis of Erraji, Armin
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas
Acoustic diffraction off and around geometrical obstacles is an important phenomenon that occurs if the compared wave length is not small or big against the geometrical dimensions. Simulating acoustic diffraction is usually done in the time or frequency domain using simulation methods that solve the wave equation, i.e. naturally take this effect into account. Calculating a sound field this way results in a high spatial resolution but is yet so computationally expensive, that it is not suitible for almost any nonfundamental problem.
Acoustic simulations using the Geometrical Acoustics principle, on the other hand, are able to rapidly generate sound transmission information for distinct solitaire positions of a source and a receiver, which makes this approach feasible for realtime auralization.
In this master thesis, acoustic sound diffraction shall be investigated using geometrical techniques.

Direction of Arrival Estimation of Early Reflections Using Compressive Sensing
Bachelor Thesis of Förster, Jonas
Betreuer: Berzborn, Marco / Berzborn, Marco
Beamforming techniques are nowadays commonly used to detect the direction of arrival
(DOA) of sound waves arriving at spherical microphone arrays (SMAs). The conventional
planewave decomposition beamformer(PWDBF) can be used to estimate the DOA with a
maximal directivity function solving a ?2  minimisation problem. However, since the number
of microphones in SMAs is limited by physical constraints, the PWDBF method suffers from
low spatial resolution in the practical use case.
The spatial resolution can be improved by applying compressive sensing, the socalled compress
ive beamforming(CB). The main requirement to apply compressive beamforming is sparsity in
the solution of the problem. That means that the minimisation problem is underdetermined
and can be solved with ?1  minimisation.
In reverberant room acoustic scenarios, the condition of sparsity is not fulfilled due to meas
urement noise and too many incoming reflections. Therefore, subspacebased preprocessing
methods are used to divide the signals into two parts. The first part is assumed to consist of
the direct sound and the early reflections whereas the second part includes late reverberation
and measurement noise. The first part is then assumed to be sparse and the directions of
arrival of the direct sound and the early reflections can be estimated using CB.
In this work, the performance of CB with and without the subspacebased preprocessing
methods is compared with the performance of PWDBF and MUltiple SIgnal Classifica
tion(MUSIC). This is done in two simulation scenarios. In the first scenario, planewave
sources are generated analytically and measurement noise is added. In the second, the methods
are applied on directional room impulse responses.
The focus of the analyses is on the influence of measurement noise and late reverberation to
the simulations with respect to their effects on sparsity of the problem and the estimation of
the DOA of the primary sources and reflections.

Dynamic CrosstalkCancellation with Room Compensation for Immersive CAVEEnvironments
Master Thesis of Röcher, Eric
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Kohnen, Michael / Stienen, Jonas
Auralization of virtual scenes using binaural synthesis enables a realistic reproduction of auditory events, which intensifies the feeling of immersion in Virtual Reality applications. Headphones are usually used with binaural synthesis; however, they tend to constrain the users immersion. To retain the immersion, a speaker setup with a proceeding CrosstalkCancellation filter can be used. In large CAVEVRSystems with extensive acoustically hard projection surfaces, which partly or completely surround the user, challenging acoustical conditions are to be expected. Not only is the speaker’s placement limited to positions above the hard projection surfaces but furthermore, due to the nonabsorbent surfaces, distinct early reflections superimpose the useful signal. This results in a not insignificant change in the overall reverberation time. Therefore a CTC filter design is proposed which compensates those early reflections. Based on the simple room geometry in CAVEEnvironments, sound transmission paths can be easily estimated. These estimations can then be used to integrate room characteristics into the design of a dynamic CTCsystem and lead to an improved playback quality for rooms with challenging acoustical conditions.
The proposed procedure will be instrumentally reviewed for its performance in the aixCAVE, the VREnvironment of the RWTH Aachen University. The tests will focus on accuracy and introduced errors for each individual hardware as well as software component of the system. Inherent System latencies and positional inaccuracies like for example the CAVE’s tracking device and transmission path delays will be taken into consideration.

Dynamische Modenkompensation in Quaderräumen
Bachelor Thesis of Voth, Markus
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Kohnen, Michael / Stienen, Jonas
In small rooms, especially with parallel and perfect reflecting walls, there occur distinct standing waves. Due to standing waves, there are constructive and destructive
interferences for single, deep frequencies. In respect of the position, this inhomogeneous spread of sound pressure is perceived by the listener as an unbalanced
an unpleasant soundfield. Especially in the aixCAVE the causes for standing waves exist, as mentioned above. The only possible optimization is a software solution,
because walls and floor are used for rear projections.
Up to now, only static software solutions are considered for the compensation of room and transducer effects. So in this bachelor thesis it is explored, how
far the spectral course of sound pressure level may be equalized in cuboid rooms with a listening position adaptive compensation of modes. Therefore an analytical
frequency response estimation and an adaptive IIRfilter concept with adapted sampling rate is implemented. To reduce the computational complexity, also a
symmetry optimized lookuptable concept is developed. The results of the compensation are validated with suitable simulations and measurements.
In the process, the estimation of modes in the aixCAVE is proved as difficult, because of the complicated boundary conditions, appearing by an open
and absorbing ceiling. With a perfect estimation, the developed software compensation of modes in the aixCAVE provides a realistic improvement of more than 20dB.

Embedded Microphone Equalization for Array Applications
Master Thesis of Maintz, Thomas
Betreuer: Berzborn, Marco
The analysis of a sound field with an array of microphones reveals detailed information
about its directional properties. To achieve a high spatial resolution many microphones
are required, typically resulting in high costs when using standard electret microphones.
To reduce costs, Micro Electro Mechanical System (MEMS) microphones are used.
These, however, show a nonideal frequency response. Thus, the frequency and phase
response of all individual microphones have to be equalized for their application in
microphone arrays.
The internal logic of Field Programmable Gate Arrays (FPGA) can be customized to
implement a liveequalization of microphones. In this thesis the Xilinx Zynq FPGA
is choosen for this purpose. In addition to the programmable logic block the Xilinx
Zynq platform includes two ARMprocessors which connect to the programmable logic
block via standardized protocols. These can be used to interface the FPGA logic and
program the filter coefficients into the programmable logic. Further, optimized atomic
architectures for digital signal processing are provided within the Zynq FPGA.
This thesis outlines the conception and implementation of a scalable convolution kernel
with minimum latency for the equalization of larger scale microphone arrays in hardware.
The implementation is carried out on the Xilinx Zynq FPGA platform relying on pipelining.
Parallel FIR and BiquadFilters are implemented and can be customized for each
microphone channel individually. By applying a hybrid convolution technique with time
and frequency domain convolution, the length of the equalization filters can be extended.
To reduce the total number of FFT operations the symmetry properties in the spectra of
real valued time signals are used. The solution will be used inside a spherical microphone
array yielding a liveequalization of the individual microphone signals. Eventually, the
output signals are converted to a MADI stream allowing for a standardized interface to
a measurement PC. Further, the implementation is analyzed with respect to the limitations
of the used fixedpoint arithmetic. Finally, results are compared to computational
results of a Golden Reference Modell implemented in floatingpoint arithmetic.

English: Physicsbased realtime auralization with the game and VR environment ‘Unity’
Bachelor Thesis of Andreas, Maurice
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas
For the creation of virtual environments not only technical and artistic abilities are needed. Without the use of StateoftheArtSoftware the enormous effort of modeling, scripting and developing individual extension scripts is not viable. The highly specialized development environment \grqq Unity\grqq{} by Unity Technologies Inc. offers an extensive base for easy implementation not only for the gaming industry but also for Virtual Reality. For this purpose integrable modules are ready to use a number of VRDevices like the Oculus Rift, LLC or the \grqq HTC Vive\grqq{}. Furthermore Unity offers extensive possibilities to program plugins and gather objects of the virtual environment with scripts written in C\#. This bachelor thesis deals with the creation of a connection between Unity and VA. This enables the usage of physically based realtime auralization and all other available reproduction methods in combination with VR glasses. The result will be a fullfledged multimodal VR system which can be used to demonstrate interactively environment noise or for building and room acoustic studies. On top of that it can be used for studies about multi modal perception which have to be executed under controllable and reproducible conditions.

Estimation of room geometry based on impulse responses
Bachelor Thesis of Maintz, Thomas
Forschungsgebiet: Akustische Virtuelle Realität / Raum und Bauakustik
Betreuer: Aspöck, Lukas
Virtual Reality methods are used to create an immersive environment where the user is able to freely interact. Besides visual perception acoustic perception is of high importance. The process of auralization makes scenes audible and aligns the acoustic and visual feedback. For this procedure simulated room impulse responses are required which represent the room's acoustical characteristics. Room impulse responses contain early reflections within the first 50 ms. Those reflections occur on bounding surfaces and appear as peaks in room impulse responses. These peaks can be approximated by the image source model which is based on the principles of geometric acoustics. An algorithm was developed which calculates the bounding surfaces and in this way estimates the room geometry. No a priori knowledge is required except for the speed of sound and the constellation of receivers. Since the identification of higher order image sources is not trivial, mainly convex room geometries are investigated. In these spaces independent of the source and the receiver positions all first order image sources are audible. Peaks of the room impulse response are converted into estimated propagation paths using methods of Euclidean Distance Matrix and Multidimensional Scaling. Based on estimated positions of the original source and image sources bounding surfaces are calculated. In order to validate the result of the estimates, a measure of the error is introduced taking into account acoustic and geometric deviations.

Evaluation of audio signal synthesis and network transmission for realtime auralization in Virtual Reality
Bachelor Thesis of Heimes, Anne
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas
In virtual reality applications, technical challenges are the biggest obstacle when producing a virtual scene in a quality standard, which achieves a satisfactory immersion. For this purpose, various components are integrated into an overall system, which separates the tasks and divides the burden of complicated calculations. Typical examples are CAVE systems, which usually use dedicated computers for visualization, tracking and auralization. They communicate exclusively via network interfaces. Extensions, which require a correspondingly high calculation effort, can be integrated via the acquisition of additional computing units without slowing the system down. An important aspect of interactive VR applications is the systems reaction to user action. In sound propagation simulation, these changes lead to new calculation of the parameters for the digital signal processing, which imprints the effects of the propagation on an input signal. These input signals of the realtime processing are usually read out directly from the main memory without significant expenditure. A different situation occurs if the input signal is not directly available, but must be generated from an artificial parameterizable calculation model or a complex physical model by means of modal analysis and synthesis. In this case, mathematical operations are carried out that additionally burden the processor and thus compete for resources. In this bachelor thesis, the aim is to examine under which preconditions the generation of signals from a virtual sound source on a dedicated computer and the subsequent transmission to the auralization computer via a network interface are reasonable. System components and complexity of signal generation as well as latency and transmission rates are taken into account in order to formulate a general recommendation.

Extention of a Model of Open Pit Mines for Optimization of Noise Emissions
Master Thesis of Uber, Thomas
Betreuer: Vorländer, Michael / Aspöck, Lukas

Filter Design for Sound Insulation Auralization
Master Thesis of Heimes, Anne
Forschungsgebiet: Raum und Bauakustik
Betreuer: Imran, Muhammad
Recent years’ surveys revealed that there is a negative influence of noise disturbances on performance of the humans in private as well as commercial working sites. These noise disturbances are present within the built environments and/or might be from outdoor sources. An extensive research is carried out to estimate the sound propagation and transmission in buildings. Different methods are available for auralization of sound insulation between connected rooms in compliance with the standardized data formats of sound insulation and building structure geometries. However, there still exist certain challenges to be addressed to accurately construct transfer functions between source and receiver rooms.
Several simplifications exist in available building acoustic auralization research. Some simplifications are implicit in the formulation on which the EN 12354 is based. In first place, the incident sound pressure is considered to be equal for all transmission paths. Similarly, the same incident sound power hits all elements, independently of the source position and room geometry. Additionally, influence of the source room reverberation, the directionality of the sound source, and the ratio between direct and reverberant energy inside the source room are integral part of transfer functions. Secondly, the transfer functions calculated from source room to receiver room are only valid for point to point transmission, however, the extended walls are always present in real situations. In the receiving room, the simplification is made that the sound is apparently radiated from one point representing the whole bending wave pattern on the wall.
This research will focus on addressing these challenges for physically correct representation of building acoustics auralization in virtual reality integrated with 3D audiovisual technology. The walls should be considered as plane sources and bending wave patterns will be addressed in order to be able to properly construct the transfer function with correct phase information. A room acoustical simulation is necessary for generating final transfer function between walls and the receiver rather than using a measured impulse response, so that the geometries and absorptions might be fit to the properties desired by the user of the auralization. In addition the receiver room transfer function from radiating walls to the receiver are required to be designed in such a way that not only indoor sound sources will be addressed but the outdoor moving sources will also be handled. In this way a physically more accurate building acoustic auralization framework will be developed in 3D visual technology and different psychoacoustic experiments will be possible in virtually reality for evaluation of noise and comfort.

Filter design of diffraction in auralization of urban environments
Bachelor Thesis of Filbert, Daniel
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas
The acoustic propagation of sound in the open field or in the interior can be calculated effectively and in high quality by geometric simulation methods. However, phenomena that are mainly due to wavebased properties are implemented insufficiently or only with comparatively complex calculation models by these approaches.
In this bachelor theses, wellknown methods for the filter design of diffraction simulation in geometric acoustics will are investigated and applied for the auralization of urban environments, in which mainly wave fronts of low order come to account.

Filteroptimisation for 3Daudio reproduction
Bachelor Thesis of Klein, Simon
Forschungsgebiet: Psychoakustik
Betreuer: Kohnen, Michael / Stienen, Jonas
Nowadays, headmounted displays (HMD) like the Oculus Rift or the HTC Vive allow an immersive experience of virtual environments even for the consumer market. To present a correct audio stimulus beside these visual reproductions loudspeakerbased reproduction methods can be used. In this thesis such loudspeakerbased reproduction method should be optimized in terms of computational efficiency by investigating the possibilities of simplifying or shortening the used filters. Furthermore the results of such an optimization will be investigation in terms of subjective audibility (i.e. a short listening experiment) and objective parameters (e.g. change in energy of the filters). Both the optimization and listening test can be coded in MATLAB using the ITAToolbox.
Requirements
 Programming skills (MATLAB)
 Work independendly
 Interest in the topic
 Basics of Acoustics (e.g. ‘Einführung in die Akustik’)

How much does the sound field in auditoria change from one position to the next?
Bachelor Thesis of Hasti, Henry
Forschungsgebiet: Raum und Bauakustik / Akustische Messtechnik
Betreuer: Witew, Ingo
Room acoustical measurements with microphone arrays have shown that the sound field in auditoria changes significantly from one position to the next. This gives rise to the question how valid acoustical measurements in architectural acoustics are. This question that is of core relevance when it comes to characterizing the sound field in rooms based on a small number or singular measurements.
At the Institute of Technical Acoustics, a measurement robot was designed, capable of conducting automated highresolution soundfield measurements in auditoria over a larger area (5.5 x 8.0 m). This data can be used to investigate how severely the sound field changes from one position to the next.
In this thesis acoustical measurements are to be conducted in different auditoria (lecture rooms and concert halls) to collect data that can serve as the foundation to derive a relationship between the change in acoustic conditions an the distance between two measurement positions. Goal of this thesis will be to answer two important questions:
• For what size a region is a single measurement in a room valid?
• How accurate need measurements to be documented so that the results are reproducible?
This knowledge is of special interest to make acoustical measurements in architectural acoustics more efficient.

Implementation of VBAP and DBAP with Evaluation using Measurements and Binaural Simulation Model
Bachelor Thesis of Bassiri, Sina
Betreuer: Vorländer, Michael / Kohnen, Michael
To create an acoustic virtual reality with the best possible realistic perception of sound
sources for a listener in a closed space, it is necessary to analyse the spatial sound in certain
areas. In this thesis two methods are considered: Vector Base Amplitude Panning (VBAP)
and DistanceBased Amplitude Panning (DBAP). Thesc techniques arc able to create virtual
sound sources for a listener by controlling several speakers.
Important cues like the Interaural Time Difference (ITD) and the Interaural Level Difference
(ILD) together with the frequency spectrum are analysed and evaluated. This aspects will be
investigated by looking at the head direction and at the movement of the virtual source.
The aim of this bachelor thesis is to implement the methods VBAP and DBAP in MatLab
and to write a corresponding test script which plays different sounds with different methods
and sound source locations. These sounds are recorded by a dummy head and they are
evaluated in terms of ITD, ILD and frequency spectrum within the test script. With existing
recordings of Head Related Transfer Functions (HRTFs) of the dummy head the evaluation
is accomplished. In addition there will be a comparison with an auditory model by Dietz,
which represents a model of the human hearing.
The dependence of the speaker positions and position in the head related coordinatesystem
are very important for the perception of the sound source. The results for DBAP's evaluation
were unfavourable. However, VBAP works for the greater part very solid.

InSitu determination of room acoustic boundary conditions based on invese simulation techniques
Master Thesis of Knauber, Fabian
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Pelzer, Sönke / Vorländer, Michael
The acoustical transformation from a real existing room into a congruent simulation model requires
a precise room model replication and appropriate wall boundary conditions, which replicate the wall
material distribution of this room. The room model is easy to build, whereas the wall boundaries
are challenging to measure insitu. Typical room acoustic simulation software uses a room model
and the absorption coefficients to calculate an impulse response. While trying to find the best
possible match of the simulation result to a given impulse response measurement, certain absorption
coefficients have to be set. This problem can not be analytically solved, but it can be formulated as
a nonlinear minimization problem, which can be iteratively solved using optimization algorithms.
In this work a MATLAB tool was developed that implements an absorption coefficient determination,
by using the optimization toolbox together with a new method for the fast generation of energy
decay curves from a given set of absorption coefficients. Furthermore, the optimization was improved
by formulating an acoustically reasonable cost function. The implementation was validated versus
simulated rooms as well as applied to measurements of different rooms.

Investigation of Thermal and Nonlinear Distortions in Small Loudspeakers
Master Thesis of Mecking, Jens
Forschungsgebiet: Elektroakustik
Betreuer: Behler, Gottfried / MüllerTrapet, Markus
This thesis deals with the characterization and modeling of small dynamic loudspeakers which are relevant due to their use in mobile devices (smart phones, tablet computers, etc.). These socalled microspeakers show considerable deviations from the idealized model of a linear timeinvariant system which allows its electrical, mechanical and acoustical characterization by means of a time independent transfer function. Although the reasons for these deviations are wellknown for the case of large dynamic loudspeakers the transferability of the results to the case of microspeakers is questionable due to the altered mechanical composition concerning structure, material and size. An exact description of the occuring distortions would eventually allow their compensation and thus lead to an improved playback quality.
During operation the temperature of the system rises which leads to changes in the electrical and mechanical properties and therefore also in the linear transfer properties of the transducer. An uncontrolled temperature increase can cause mechanical damage and eventually even the destruction of the device. An accurate description of the speaker’s thermal behavior is therefore necessary to compensate these undesired effects. In order to carry out thermal investigations on the system, a precise knowledge of the device temperature is essential. Therefore, a method to measure the temperature via the DC resistance of the voice coil was validated regarding accuracy and reproducibility. This method was subsequently used to characterize a loudspeaker in the framework of thermal models which predict the temperature as a function of the electical input power. Furthermore, the temperature dependence of the linear ThieleSmall parameters was investigated.
In addition to thermal effects, microspeakers show nonlinear distortions in their transfer characteristic if operated in the largesignal domain. The HarmonicBalance method was used in order to characterize the loudspeaker in the framework of an extended nonlinear model. Therefore, generalized transfer functions for higher orders of the input signal were derived and fitted to measurement data obtained in the nonlinear regime. A special focus was laid on the integration of the linear loudspeaker parameters into the nonlinear model.

Investigation on headphone reproduction using individual and nonindividual HRTFs
Master Thesis of Reffgen, Matthias
Betreuer: Vorländer, Michael
The use of binaural synthesis to generate virtual aeoustieal environments for numerous
eontexts is nowadays of peeuliar interest. Many studies foeus on the preeise
determination of individual headrelated transfer functions (HRTF ) , the investigation
of anthropometrie data as well as its influenee in time and frequeney domain.
Alternatively different proeedures based on simplified geometrieal models approximate
speeifie indiviudual HRTFs to reduee the measurement effort related to
the determination of individual HRTFs. This seems to be essentially for the eommereial
use of binaural teehnology.
To evaluate the performance of different individualizations, well known and new
investigation methods were implemented to be used together with a headmounted
display and applied in several listening tests. Along with the employment of individualization
teehniques, such as headphone equalization or ITDadjustment, also
the influence of additionally presented visual stimuli and possible applications of
headmounted displays in listening tests were discussed.
The results of these investigations should be used to evaluate possible applications
of binaural technology in nonscientific multimedia contexts.

Listening test on acoustic immersion in virtual reality
Master Thesis of Jarmer, Fabian
Forschungsgebiet: Akustische Virtuelle Realität / Psychoakustik
Betreuer: Kohnen, Michael / Aspöck, Lukas

Listening test on acoustic immersion in virtual reality
Bachelor Thesis of Lian Esthefany Gomez De Pasquale
Forschungsgebiet: / Akustische Virtuelle Realität
Betreuer: Aspöck, Lukas
A realistic virtual reproduction of a room acoustic scene deeply relies on the accuracy of the used
room model and the specified boundary conditions, in particular the absorption coefficients. The
determination of the absorption coefficients is usually a challenge, in spite of the multiple available
methods for its obtaining. A standardized method estimates the absorption coefficient of random
incidence of a material's sample by performing an acoustic measurement when placing it in a
reverberation chamber, and comparing the results with those of another measurement of the empty
chamber. As this method has some drawbacks and limitations, innovative approaches have been
proposed that use an acoustic simulation of the room and modify the input parameters until a
match is found for an acoustic measurement of the room of the identical situation.
This work presents, analyzes and compares the results obtained through the methods described
when they are applied to different sample layouts, in shape and size, both in diffuse and nondiffuse
sound fields. In addition, the accuracy of the results obtained through these methods is tested by
altering the input parameters of the model of the simulated scenarios.

Measurement Uncertainties in Vibroacoustic Problems with multiple Degrees of Freedom
Master Thesis of Dreier, Christian
Forschungsgebiet: Transferpfadanalyse / synthese
Betreuer: MüllerGiebeler, Mark / Vorländer, Michael
In contrast to sound sources in fluids and especially air there is a lack of a generally
accepted characterization method of structureborne sound sources causing considerable
difficulties in many technical applications. Industrial development chains
with modular development stages need precise source descriptions concerning their
physical behaviour. In the automotive sector for example the vibration behaviour
of an engine coupled with its subframe and the vehicle body, can be used to predict
the interior sound of a vehicle. A precise look on the example of a drivetrain
reveals this task not being trivial: the resulting interior sound field is a coupled
inter action between the direct air borne sound radiation of the engine with an
indirect structureborne sound radiation due to vibroacoustic phenomena.
Even though the topic of structureborne sound source characterization has been
questioned several years  like on transfer path analysis (TPA)  until now no
measurement device solving these problems exists (see chapter 3). Furthermore,
the concept of Frequency Response Functions ( FRF )  which is used to describe the
dynamic behaviour of a structural model between an applied load and its resulting
vibration  is characterised in nonparametric way using a mobility matrix.
However, employment of modern measurement setups is limited in terms of suffi.cient
precision by sensing only nine elements of a sixbysix matrix (for single point
excitation) in the frequency range of interest, the audible spectrum (see chapter
2.3). In reality up to today no precise moment um exciter exists so that inherent
measurement uncertainties in the conventional TPA occur. In order to deal with
this severe vibroacoustic measurement problem, the utilization of F initeElement
Method (FEM)  which has blossomed out to a competitive alternative due to increased
computation ability  is able to tremendously expand the scope of analyses
concerning structuralacoustical coupling (see chapters 4.4 and 5) by enabling a
distinct activation of every single DOF as excitation source.
Modern approaches in research and development of industrial applications usually
are based on elaborated advance developments, using simulations in order to predict . .
the properties of a future product as precise as possible. For example, in a fully
CADdriven physical simulation of an electric drivetrain, the machanical impedance
distribution on the engine core should be auralized afterwards to virtually predict
the interior vehicle sound.
In contrast to the conventionally measured TDOF (see chapter 6 the complex
movement of the electric engine not only excites translational velocities as conventionally considered, exhibiting the question of the audibility of rotational degrees
of freedom ( RDOF) .
The concern of this thesis is to deal with the multidimensional measurement problem
in vibroacoustics by using numerical simulation techniques and to make the
findings metrologically usable. Finally, it is proposed an approach for measuring
RDOF with conventional measurement equipment to deduce the importance of
rotational degrees of freedom in vibroacoustic transfer paths.

Method for the assessment of simulated electric car drives by means of psychoacoustic parameters
Bachelor Thesis of Vermeulen, Markus
Betreuer: / Vorländer, Michael

Modeling binaural receivers for hybrid acoustic simulations
Forschungsgebiet: Akustische Virtuelle Realität / Binauraltechnik
Betreuer: Schäfer, Philipp
Hybrid acoustic simulation tools combine wavebased (FEM/BEM) and geometrical acoustic (ray tracing) methods in order to balance computational effort and accuracy. In this approach, lower and medium frequencies  where wavebased effects are dominant  are simulated using the former while higher frequencies are simulated using geometrical methods to save computation time.
An important aspect of such a simulation is a binaural receiver that allows perceiving sound from different directions. For the wavebased part, this can be achieved including a geometry of e.g. an artificial head in the scene. Since at lower frequencies finer details of this geometry do not affect the wavefield while increasing the computational effort, a simplification of this geometry seems reasonable.
Thus, different geometric models for a binaural receiver are to be compared in this thesis using wavebased simulation methods. The performance  regarding accuracy of the results and computational effort  is to be investigated in respect to the simulated frequency range.
Requirements
 Basic knowledge of binaural hearing
 Understanding of acoustic fundamentals
 Working selfreliantly
 Interest in numerics and acoustics
Optional skills
 Knowledge of FEM / BEM
 Knowledge of CAD modeling
 Basic Matlab skills

Modellierung und Evaluierung der Schallabstrahlung elektrischer Antriebe
Forschungsgebiet: Maschinenakustik und Diagnose und Transpferpfadanalyse / Numerische Akustik
Betreuer: MüllerGiebeler, Mark

Modelling of moving sound sources using compressive beamforming for auralization
Master Thesis of Li, Yan
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Meng, Fanyu
Compressive beamforming (CB) is a method to reconstruct sparse signals using few measurements by solving a convex minimization problem. CB not only has good localization resolution, but is also able to reconstruct source signals with phase information. For moving sound sources, CB guarantees an acceptable localization resolution; Besides, phase information benefits auralization by completely reconstructing source signals. This paper focuses on obtaining the signals of moving sound sources for auralization using CB. A microphone array will be applied to record a passby moving sound source. The deployment of the array should be determined in terms of a better performance. Subsequently, the recordings are processed by CB to localize and reconstruct the source signal. Finally, the reconstructed signal will be added to the moving sound source in virtual reality (VR) for auralization.

Optimization of a sound insulation test bench regarding its measurement accuracy
Bachelor Thesis of Mattern, ArneHeinz
Forschungsgebiet: Raum und Bauakustik / Akustische Messtechnik
Betreuer: Mecking, Jens
One of the most important measures of sound insulation is the sound reduction index.
It describes the property of a component regarding sound transmission as a function of frequency and is therefore decisive for the sound propagation in buildings. The sound reduction index is determined through measurements in a test bench according to EN ISO 10140. An important aspect that must be considered during measurements is the flanking transmission, i.e. the transmission of power through objects that are flanking the tested component. This additional power increases the measured sound pressure in the receiver room and thus the measured sound reduction index is lower than its true value. But, because the sound reduction index is supposed to be the property of a component and not dependent on the room itself, the flanking transmission must be suppressed as far as possible.
The aim of this thesis is to improve the measurement accuracy of the test bench, which is primarily used during the acoustic lab course. For this purpose, a characterization of the test bench is achieved by comparing the measured and calculated values for the sound reduction index. Then the possible flanking paths are characterized and analyzed. Based on these results constructive solutions are derived and implemented. In the end, the proposed solutions should be evaluated. A beforeandafter comparison of the measuring accuracy is suitable for this purpose.

Optimized sound propagation simulation for auralization of outdoor noise in urban environments
Bachelor Thesis of Jansen, Julian
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas
Noise maps are key tools of noise protection in urban environments. The expected, averaged noise exposure is visualized for every location within the surveyed area. This visual representation, however, does not allow a statement about the characteristics of the perceived sounds.
The aim of this thesis is a realtime simulation of the sound propagation from various sound sources to a receiver in a threedimensional city model, to make an auralisation of the perceived sounds possible.
The size of the city model datasets requires the implementation of optimization approaches, which limit the number of calculated propagation paths and geometric objects, if realtime calculation is to be achieved. Crucial to this is an apriori analysis of the source signal properties.

Parallelisierung einer EchtzeitRaumakustikSimulation für Mehrkernprozessoren
Master Thesis of Schallenberg, Ralf
Betreuer: Pelzer, Sönke / Vorländer, Michael

Perception of Spherical Harmonics Source Directivity Patterns in Rooms
Bachelor Thesis of Wolf, Gregor
Betreuer: Klein, Johannes / Berzborn, Marco
Room acoustical measurements according to the international standard ISO 3382 are executed
with the requirement of directivity of source and receiver to be omnidirectional to ensure
the comparability of standardized parameters. Due to this, directivity patterns of ordinary
sources, like human speakers or instruments, are not respected in these measurements. For
realistic auralization and precise room acoustical analysis the measurement of room impulse
responses including the desired directivity is required.
Due to practical considerations, every room impulse measurement is limited in terms of the
spatial resolution in spherical harmonics. In this regard, the question of the human perception
accuracy of the spatial resolution arises. Listening tests have proven, that differences between
strongly directive sources and rather omnidirectional sources are audible.
The goal of this thesis is to evaluate the human perception accuracy of sound directivity
pattern in rooms with consideration of the spatial resolution of the directivity patterns. For
this task, defined directivity patterns of different spatial resolutions are simulated in different
scene models with the objective to create signals for a listening test.
Based on these simulations a listening test of the concluded sources and scenes is implemented
and executed. With the data it is possible to evaluate how detailed the perception is and
to analyse a possible threshold of audibility regarding the spatial resolution of the different
sources.
.

Psychoakustische Untersuchung der Geräuschqualität von Elektrofahrzeugen
Forschungsgebiet: Maschinenakustik und Diagnose und Transpferpfadanalyse / Psychoakustik
Betreuer: MüllerGiebeler, Mark

Realtime auralization of numerous sound sources
Master Thesis of Mösch, Lucas
Forschungsgebiet: Akustische Virtuelle Realität
Betreuer: Stienen, Jonas / Aspöck, Lukas
Simultaneous rendering of numerous sound sources in real time still poses a challenge today. Especially for rendering virtual environments, where up to hundreds or thousands of potential sound source objects are present, not enough processing power is available to provide a convincing aural experience with common auralization approaches. This thesis proposes a new auralization pipeline which is focused on real time processing, while maintaining the plausibility of the virtual scene. This is done via stateoftheart culling and clustering algorithms that reduce the number and complexity of sound source calculations, as well as a sophisticated sound source and cluster transitioning method, that ensures a seamless audio feedback for the user.

Sound source localization on a car body based on beamforming using transfer functions
Master Thesis of Havolli, Albulena
Betreuer: Nau, Clemens / Berzborn, Marco
The acoustic and vibroacoustictechnical comfort of vehicles (NVH) raises in importance, especially in the premium segment. In order to fully exploit its own development potential and continue to make progress towards its competitors, the advancement of acoustic development tools is of central importance for OEM (Original Equipment Manufacturer). As a successful development tool for the localization and characterization of noise outside and inside the vehicle, the beamforming has proven to be the method. The localization of a sound source which inducts structureborne sound into the car body, which is emitted as an air sound into the vehicle cabin by the interior boundary, is currently not possible. Thus, with the existing conventional beamforming algorithms (CBF), only a locating of the airborne radiation, but not of the structureborne sound generation, is possible.
This master thesis is intended to contribute to the improvement of beamforming for sound source detection in the vehicle interior by using transfer functions. For this purpose, in the first step, the transfer functions of defined structural points of the car body are determined experimentally (by striking with the impact hammer). In the next step, these transfer functions are implemented in an existing beamforming environment.
To validate the beamforming approach using transfer functions, the car body is excited with a shaker at the defined positions. The localization results of this approach and the corresponding method without the inclusion of transfer functions are recorded, compared and qualitatively evaluated.