Binaural simulation is the recreation of a three-dimensional audio environment around a listener's head. The binaural simulation of structural acoustic data would open new opportunities in virtual prototyping and simulation. By modeling the structure as an array of vibrating monopoles and applying Head Related Transfer Functions (HRTFs) to each of the sources, a binaural simulation of this type can be created. Unfortunately, this simulation method requires an extensive amount of computer power and speed for real-time simulation, more so than is available with current technology. The objective of this research is to reduce the number of computations required in the binaural simulation of structural acoustic data. This thesis details two methods of reducing the number of real-time calculations required in this binaural analysis: singular value decomposition (SVD), and equivalent source reduction (ESR). The SVD method reduces the complexity of the HRTF computations by breaking the HRTFs into dominant singular values and vectors. The ESR method reduces the number of sources to be analyzed in real-time by replacing sources on the scale of a structural wavelength with sources on the scale of an acoustic wavelength. The ESR and SVD reduction methods can be combined to provide an estimated computation time reduction of 99.4%. In addition, preliminary tests show that there is a 97% correlation between the results of the combined reduction methods and the results found with current binaural simulation techniques. / Master of Science

Virtual acoustic techniques can be used to create virtual listening environments for multiple purposes. Using multi-speaker reproduction, a physical environment can take on the acoustical appearance of another environment. Implementation of this environment auralization could change the way customers evaluate speakers in a retail store. The objective of this research is to develop a virtual acoustic showroom using a multi- speaker system. The two main components to the virtual acoustic showroom are simulating living environments using the image source method, and simulating speaker responses using inverse filtering. The image source method is used to simulate realistic living environments by filtering the environment impulse response by frequency-dependant absorption coefficients of typical building materials. Psychoacoustic tests show that listeners can match virtual acoustic cues with appropriate virtual visual cues. Inverse filtering is used to "replace" the frequency response function of one speaker with another, allowing a single set of speakers to represent any number of other speakers. Psychoacoustic tests show that listeners could not distinguish the difference between the original speaker and the reference speaker that was mimicking the original. The two components of this system are shown to be accurate both empirically and psychologically. / Master of Science

The numerical simulation of large-scale virtual acoustics and physical modelling synthesis is a computationally expensive process. Time stepping methods, such as finite difference time domain, can be used to simulate wave behaviour in models of three-dimensional room acoustics and virtual instruments. In the absence of any form of simplifying assumptions, and at high audio sample rates, this can lead to simulations that require many hours of computation on a standard Central Processing Unit (CPU). In recent years the video game industry has driven the development of Graphics Processing Units (GPUs) that are now capable of multi-teraflop performance using highly parallel architectures. Whilst these devices are primarily designed for graphics calculations, they can also be used for general purpose computing. This thesis explores the use of such hardware to accelerate simulations of three-dimensional acoustic wave propagation, and embedded systems that create physical models for the synthesis of sound. Test case simulations of virtual acoustics are used to compare the performance of workstation CPUs to that of Nvidia’s Tesla GPU hardware. Using representative multicore CPU benchmarks, such simulations can be accelerated in the order of 5X for single precision and 3X for double precision floating-point arithmetic. Optimisation strategies are examined for maximising GPU performance when using single devices, as well as for multiple device codes that can compute simulations using billions of grid points. This allows the simulation of room models of several thousand cubic metres at audio rates such as 44.1kHz, all within a useable time scale. The performance of alternative finite difference schemes is explored, as well as strategies for the efficient implementation of boundary conditions. Creating physical models of acoustic instruments requires embedded systems that often rely on sparse linear algebra operations. The performance efficiency of various sparse matrix storage formats is detailed in terms of the fundamental operations that are required to compute complex models, with an optimised storage system achieving substantial performance gains over more generalised formats. An integrated instrument model of the timpani drum is used to demonstrate the performance gains that are possible using the optimisation strategies developed through this thesis.

Este trabalho apresenta diversas propostas associadas ao uso ótimo de funções de transferência relativas à cabeça (HRTFs) em sistemas de reprodução acústica virtual por fones de ouvido. Estas propostas permitem personalizar a HRTF a indivíduos particulares, tomando como base uma combinação da modelagem estrutural e morfológica de HRTFs. Dentro do contexto da modelagem estrutural, o presente trabalho se concentrou no estudo da contribuição do pinna à HRTF. O pinna é a estrutura anatômica responsável pela percepção de elevação. Assim, o primeiro passo foi extrair um conjunto de funções de transferência relativas ao pinna (PRTFs) das HRTFs de uma base de dados. Para tanto, foram usadas diversas técnicas como análise preditiva linear para rastrear as ressonâncias, janelamento para eliminar a influência do torso, funções de autocorrelação e de atraso de grupo para salientar as antirressonâncias, e outros algoritmos para combinar ressonâncias e antirressonâncias em apenas uma magnitude espectral. Usando essa nova base de dados de PRTFs e parâmetros antropométricos propostos mais outros registrados na base de dados, um espaço vetorial correspondente à antropometria do pinna foi mapeado linearmente em um espaço vetorial correspondente às características espectrais da PRTF, calculando-se assim várias transformações lineares para estimação de novas PRTFs fora da base de dados. A estimação atingiu 66% de reconstrução no grupo de treino. O trabalho está orientado à exploração das características espectrais importantes na percepção de elevação, portanto, está limitado ao plano médio do hemisfério frontal, onde não existem diferenças interaurais significativas nem efeitos difrativos da cabeça. Finalmente é proposto um sistema de testes de localização de fonte sonora para validar o modelo. / This work presents several proposals associated with the optimal use of head-related transfer functions (HRTF) in virtual auditory spaces presented via headphones. These proposals lead to personalization of the HRTF to particular individuals, using a combination of the structural and morphological modeling techniques. In the context of structural modeling, this work focuses on modeling the contribution of the pinna to the HRTF. The pinna is the anatomical structure responsible for vertical sound localization. Thus, the first step was to extract a set of pinna-related transfer functions (PRTFs) from HRTFs published in a database. This was accomplished using several techniques like linear prediction analysis for tracking the resonances, windowing for eliminating the torso influence, autocorrelation and group delay functions for emphasizing the notches and other algorithms for combining resonances and notches in only one magnitude response. Using this novel database of PRTFs and a set of proposed anthropometric parameters plus some others registered in the database, a vector space corresponding to pinna anthropometry is linearly mapped into a vector space corresponding to spectral features of the PRTF, being calculated, in this way, several linear transformations for estimation of new PRTFs, outside the database. The estimation attains 66% of reconstruction in the training group. The work focuses on the exploration of spectral characteristics important for elevation perception, therefore, it is limited to the median plane where there are no meaningful interaural differences nor head diffraction effects. Finally, a system for sound localization tests is proposed in order to validate the model.

Este trabalho apresenta diversas propostas associadas ao uso ótimo de funções de transferência relativas à cabeça (HRTFs) em sistemas de reprodução acústica virtual por fones de ouvido. Estas propostas permitem personalizar a HRTF a indivíduos particulares, tomando como base uma combinação da modelagem estrutural e morfológica de HRTFs. Dentro do contexto da modelagem estrutural, o presente trabalho se concentrou no estudo da contribuição do pinna à HRTF. O pinna é a estrutura anatômica responsável pela percepção de elevação. Assim, o primeiro passo foi extrair um conjunto de funções de transferência relativas ao pinna (PRTFs) das HRTFs de uma base de dados. Para tanto, foram usadas diversas técnicas como análise preditiva linear para rastrear as ressonâncias, janelamento para eliminar a influência do torso, funções de autocorrelação e de atraso de grupo para salientar as antirressonâncias, e outros algoritmos para combinar ressonâncias e antirressonâncias em apenas uma magnitude espectral. Usando essa nova base de dados de PRTFs e parâmetros antropométricos propostos mais outros registrados na base de dados, um espaço vetorial correspondente à antropometria do pinna foi mapeado linearmente em um espaço vetorial correspondente às características espectrais da PRTF, calculando-se assim várias transformações lineares para estimação de novas PRTFs fora da base de dados. A estimação atingiu 66% de reconstrução no grupo de treino. O trabalho está orientado à exploração das características espectrais importantes na percepção de elevação, portanto, está limitado ao plano médio do hemisfério frontal, onde não existem diferenças interaurais significativas nem efeitos difrativos da cabeça. Finalmente é proposto um sistema de testes de localização de fonte sonora para validar o modelo. / This work presents several proposals associated with the optimal use of head-related transfer functions (HRTF) in virtual auditory spaces presented via headphones. These proposals lead to personalization of the HRTF to particular individuals, using a combination of the structural and morphological modeling techniques. In the context of structural modeling, this work focuses on modeling the contribution of the pinna to the HRTF. The pinna is the anatomical structure responsible for vertical sound localization. Thus, the first step was to extract a set of pinna-related transfer functions (PRTFs) from HRTFs published in a database. This was accomplished using several techniques like linear prediction analysis for tracking the resonances, windowing for eliminating the torso influence, autocorrelation and group delay functions for emphasizing the notches and other algorithms for combining resonances and notches in only one magnitude response. Using this novel database of PRTFs and a set of proposed anthropometric parameters plus some others registered in the database, a vector space corresponding to pinna anthropometry is linearly mapped into a vector space corresponding to spectral features of the PRTF, being calculated, in this way, several linear transformations for estimation of new PRTFs, outside the database. The estimation attains 66% of reconstruction in the training group. The work focuses on the exploration of spectral characteristics important for elevation perception, therefore, it is limited to the median plane where there are no meaningful interaural differences nor head diffraction effects. Finally, a system for sound localization tests is proposed in order to validate the model.

In Virtual Reality (VR) systems, the incorporation of acoustics allows for the generation of audio-visual stimuli, facilitating applications in engineering, architecture, and design. The goal of virtual acoustics is to create a realistic sound field in continuous space. Realistic virtual acoustic environments can be produced with wave-based acoustic simulations. However, rendering a sound field with a dense grid of room impulse responses (RIRs) in real-time is slow and memory-intensive. Conventionally, a more sparsely spaced grid of RIRs is used and as a workaround linear interpolation between the nearest RIRs is performed, allowing users to listen at an arbitrary location. However, the linear interpolation method reduces the quality of the sound field as it does not produce natural-sounding RIRs. The aim of this thesis is therefore to answer the question of whether we are able to achieve a better interpolation technique than linear interpolation using a machine learning approach. In this thesis, we present a novel neural network-based method for interpolating between Room Impulse Responses (RIRs). The networks were trained using RIRs from a wave-based simulation of a single 3D room and developed through a series of experiments. The experimental process was performed in three distinct stages. Firstly, we explored various representations of the RIRs: unprocessed RIRs, Short-time Fourier transform (STFT) of RIRs, and encoded STFT of the RIRs using an autoencoder. Secondly, we examined several different neural network architectures: Multi-layer perception, residual neural network, autoencoder, and U-Net. Additionally, we experimented with training the networks in a Generative Adversary Network (GAN) setting. Thirdly, we experimented with different sizes of the best-performing architecture. Results show that using an STFT representation of the RIRs combined with a residual neural network architecture yielded the most optimal results. Furthermore, we were able to outperform the established linear interpolation baseline. / Inom Virtuell Verklighet (VR) möjliggör användningen av akustik skapandet av audiovisuell stimuli, vilket underlättar tillämpningar inom ingenjörsvetenskap, arkitektur och design. Målet med virtuell akustik är att skapa ett verklighetstroget och kontinuerligt ljudfält. Verklighetstrogna virtuella akustiska miljöer kan skapas med hjälp av vågbaserade akustiska simuleringar. Men att återge ett ljudfält med ett tätt rutnät av Room Impulse Responses (RIRs) i realtid är långsamt och minneskrävande. Konventionellt används ett rutnät med glesare avstånd av RIR, och som en lösning utförs linjär interpolation mellan de närmaste RIR:erna, vilket tillåter användare att lyssna på en godtycklig plats. Den linjära interpolationen minskar dock kvaliteten på ljudfältet eftersom den inte producerar naturligt ljudande RIR:er. Syftet med detta examensarbete är därför att besvara frågan om vi kan finna en bättre interpolationsteknik än linjär interpolation med hjälp av en maskininlärningsmetod. I detta examensarbete presenterar vi en ny metod för interpolering mellan Room Impulse Responses (RIR:er) baserad på neurala nätverk. De neurala nätverken tränades med hjälp av RIR:er från en vågbaserad simulering av ett enskilt 3D-rum och utvecklades genom en serie experiment. Experimenten utfördes i tre steg. Först undersöktes olika representationer av RIR:er: obearbetade RIR:er, korttids fouriertransform (STFT) av RIR:er och kodade STFT av RIR:er med hjälp av en autoencoder. Det andra steget innefattade undersökningen av flera olika neurala nätverksarkitekturer: Multi-layer perception, residual neural network, autoencoder och U-Net. Dessutom experimenterade vi med att träna nätverken i en GAN-miljö (Generative Adversary Network). I det tredje steget experimenterade vi med olika storlekar på den mest effektiva arkitekturen. Resultaten visar att användning av en STFT-representation av RIR:er kombinerat med en residual neural nätverksarkitektur resulterade i de mest optimala resultaten. Dessutom kunde vi överträffa den etablerade linjära interpolationsbaslinjen.

[ES] Los recintos de culto son construcciones patrimoniales de importantes valores culturales, históricos, artísticos y arquitectónicos. Además, presentan una sonoridad característica que debe ser objeto de especial protección. En este sentido, la Organización de las Naciones Unidas para la Educación, la Ciencia y la Cultura (UNESCO) ha reconocido el sonido de los recintos de culto como una parte importante de nuestro patrimonio cultural inmaterial, estableciendo como prioridad su preservación. En este ámbito, la acústica virtual se presenta como una herramienta muy eficaz para el estudio del patrimonio sonoro de los recintos de culto. A través de simulaciones acústicas, es posible analizar en detalle el comportamiento acústico de los diferentes espacios y estudiar la evolución sonora de los templos. La acústica virtual permite predecir y percibir el resultado acústico de cualquier propuesta de intervención antes de su realización práctica, con lo que se presenta una nueva variable a tener en cuenta a la hora de afrontar la restauración del patrimonio: la acústica arquitectónica. Asimismo, la simulación acústica y visual ofrece la posibilidad de realizar turismo virtual por los recintos de culto, fomentando el estudio y disfrute de sus valores patrimoniales. Con todo ello, en esta investigación se ha aplicado la acústica virtual en el estudio del patrimonio sonoro, tanto interior como exterior, de la Catedral de Valencia. / [CA] Els recintes de culte són construccions patrimonials d'importants valors culturals, històrics, artístics i arquitectònics. A més, presenten una sonoritat característica que ha de ser objecte d'especial protecció. En aquest sentit, l'Organització de les Nacions Unides per a l'Educació, la Ciència i la Cultura (UNESCO) ha reconegut el so dels recintes de culte com una part important del nostre patrimoni cultural immaterial, establint com a prioritat la seua preservació. En aquest àmbit, l'acústica virtual es presenta com una eina molt eficaç per a l'estudi del patrimoni sonor dels recintes de culte. A través de simulacions acústiques, és possible analitzar detalladament el comportament acústic dels diferents espais i estudiar l'evolució sonora dels temples. L'acústica virtual permet predir i percebre el resultat acústic de qualsevol proposta d'intervenció abans de la seua realització pràctica, amb el que es presenta una nova variable a tindre en compte a l'hora d'afrontar la restauració del patrimoni: l'acústica arquitectònica. Així mateix, la simulació acústica i visual ofereix la possibilitat de realitzar turisme virtual pels recintes de culte, fomentant l'estudi i gaudi dels seus valors patrimonials. Amb tot això, en aquesta investigació s'ha aplicat l'acústica virtual en l'estudi del patrimoni sonor, tant interior com exterior, de la Catedral de València. / [EN] Places of worship are patrimonial constructions of important cultural, historical, artistic and architectural values. In addition, they have a characteristic sound that must be the object of special protection. In this sense, the United Nations Educational, Scientific and Cultural Organization (UNESCO) has recognized the sound of places of worship as an important part of our intangible cultural heritage, establishing its preservation as a priority. In this field, virtual acoustics is presented as a very effective tool for studying the sound heritage of places of worship. Through acoustic simulations, it is possible to analyze in detail the acoustic behavior of different spaces and study the sound evolution of temples. Virtual acoustics makes possible to predict and perceive the acoustic result of any intervention proposal before its practical implementation, thus presenting a new variable to consider when dealing with heritage restoration: architectural acoustics. Likewise, acoustic and visual simulation offers the possibility of carrying out virtual tourism around places of worship, promoting the study and enjoyment of their heritage values. With all this, in this research virtual acoustics has been applied in the study of the sound heritage, both inside and outside, of the Cathedral of Valencia. / Gracias al Ministerio de Ciencia e Innovación por los proyectos de investigación BIA2012-36896 y BIA2016-76957-C3-3-R en los que se enmarca este trabajo. / Díaz Rubio, E. (2023). La acústica virtual como herramienta para el estudio del patrimonio arquitectónico. Aplicación en la Catedral de Valencia [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/193837