List of Publications


DISCLAIMER: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists or to reuse any copyrighted component of this work in other works must be obtained from the publishers.



Articles in peer reviewed journals

2023

V. Vencovsky, A. Novak, O. Klimes, P. Honzik & A. Vetesnik (2023), "Distortion-product otoacoustic emissions measured using synchronized swept-sines", The Journal of the Acoustical Society of America. Vol. 153(5), pp. 2586--2586.

PDF url
Swept-sines provide a tool for fast and high-resolution measurement of evoked otoacoustic emissions. During the measurement, a response to swept-sine(s) is recorded by a probe placed in the ear canal. Otoacoustic emissions can then be extracted by various techniques, e.g., Fourier analysis, the heterodyne method, and the least-square-fitting (LSF) technique. This paper employs a technique originally proposed with exponential swept-sines, which allows for direct emission extraction from the measured intermodulation impulse response. It is shown here that the technique can be used to extract distortion-product otoacoustic emissions (DPOAEs) evoked with two simultaneous swept-sines. For proper extraction of the DPOAE phase, the technique employs previously proposed adjusted formulas for exponential swept-sines generating so-called synchronized swept-sines (SSSs). Here, the SSS technique is verified using responses derived from a numerical solution of a cochlear model and responses measured in human subjects. Although computationally much less demanding, the technique yields comparable results to those obtained by the LSF technique, which has been shown in the literature to be the most noise-robust among the emission extraction methods.
@article{vencovsky2023distortion,
    author={Vencovsky, Vaclav and Novak, Antonin and Klimes, Ondrej and Honzik, Petr and Vetesnik, Ales},
    title={Distortion-product otoacoustic emissions measured using synchronized swept-sines},
    journal={The Journal of the Acoustical Society of America},
    volume={153},
    number={5},
    doi={10.1121/10.0017976},
    url={https://doi.org/10.1121/10.0017976},
    pages={2586--2586},
    year={2023},
    publisher={AIP Publishing}
}

2022

O. Munroe, A. Novak & L. Simon (2022), "Reluctance Force Modeling and Compensation", Journal of the Audio Engineering Society. Vol. 70(3), pp. 177-184.

url
In the case of an electrodynamic loudspeaker, the reluctance force is purely a nonlinear addition to the Lorentz force. It generates second harmonic and intermodulation distortions and very low--frequency components in the force applied to the moving assembly. Being able to accurately model and then compensate this force is one of the building blocks for a feedforward system designed to linearize an electrodynamic loudspeaker. This work investigates the reluctance force formulation and proposes a more accurate model for motor structures with a shorting ring and compensation algorithm that does not require root finding or model inversions. The compensation is applied to two drive units and shown to reduce both harmonic and intermodulation distortion by around 18 dB between direct current and 1 kHz. Compensation results using parameters fitted to measured and simulated electrical impedances are compared. Finally there is a brief discussion on the implications of the results for shorting ring design.
@article{munroe2022reluctance,
    author={Munroe, Oliver and Novak, Antonin and Simon, Laurent},
    title={Reluctance Force Modeling and Compensation},
    journal={Journal of the Audio Engineering Society},
    volume={70},
    number={3},
    doi={https://doi.org/10.17743/jaes.2021.0054},
    url={https://www.aes.org/e-lib/browse.cfm?elib=21558},
    pages={177-184},
    year={2022},
    publisher={Audio Engineering Society}
}

2021

A. Novak & P. Honzik (2021), "Measurement of nonlinear distortion of MEMS microphones", Applied Acoustics. Vol. 175(0), pp. 107802.

PDF url
An experimental investigation on the nonlinear behaviour of the MEMS microphones is of rising interest as their use is growing rapidly not only in the domain of consumer audio devices but also in the measurement applications where higher precision is needed. The present work uses a recently published method allowing a distortionless harmonic excitation at high-pressure levels to measure and examine the distortion of single backplate MEMS microphones. The measured distortion is compared with the values predicted by a simple model of the nonlinearities originating in the electrostatic transduction principle. A good agreement between the predicted and the measured values of the second harmonic is shown. The level of the third harmonic is influenced by nonlinear effects not taken into account in the simple model and therefore deviates from the predicted values.
@article{Novak_2021_Mems,
    author={Novak, A. and Honzik, P.},
    title={Measurement of nonlinear distortion of MEMS microphones},
    journal={Applied Acoustics},
    volume={175},
    number={0},
    doi={https://doi.org/10.1016/j.apacoust.2020.107802},
    url={https://www.sciencedirect.com/science/article/pii/S0003682X20309075},
    pages={107802},
    year={2021},
    publisher={Elsevier}
}

L. Husnik, P. Bequin, P. Lotton & A. Novak (2021), "Digital acoustic projector-A proof of concept", Applied Acoustics. Vol. 173(0), pp. 107667.

url
Digital loudspeaker, Parametric array, Nonlinear propagation", abstract = "This paper describes a Digital Acoustic Projector (DAP), which is a hybrid sound system combining both a digital loudspeaker and a parametric array to convert directly an n-bits digital electrical signal into sound. It consists of a set of ultrasonic transducers emitting a sinusoidal signal modulated by a sampled and quantized low-frequency signal. Self demodulation effect is used to demodulate the acoustic signal during its propagation in the air. As a proof of concept, a DAP is designed to convert a 4-bit digital electrical signal into sound. First experimental results are presented.
@article{husnik173digital,
    author={Husnik, Libor and Bequin, Philippe and Lotton, Pierrick and Novak, Antonin},
    title={Digital acoustic projector-A proof of concept},
    journal={Applied Acoustics},
    volume={173},
    number={0},
    doi={https://doi.org/10.1016/j.apacoust.2020.107667},
    url={http://www.sciencedirect.com/science/article/pii/S0003682X20307714},
    pages={107667},
    year={2021},
    publisher={Elsevier}
}

2020

A. Novak, B. Lihoreau, E. Brasseur, P. Lotton & L. Simon (2020), "Guitar Pickup: Measurement of its Two-Dimensional Nonlinear Function", J. Audio Eng. Soc. Vol. 68(3), pp. 184-192.

blog post url
This paper addresses the question of why some guitar pickups distort more than others. The electromagnetic pickup of an electric guitar is a nonlinear device that can provide a pleasant distortion. Although the pickup is a simple device consisting of a coil and a few magnets or pole pieces, the measurement of its nonlinear function is a difficult task. This research shows a measurement technique that can estimate the nonlinear function of a pickup in both y and z directions of the vibrating string. The experimental results are provided for three different types of pickups: a single-coil pickup with six staggered pole-pieces, a humbucker pickup with six equal-height pole-pieces, and a humbucker rail pickup. The measured nonlinear functions of the three pick-ups are very different from each other, leading to different distortions. These experimental results confirm that the pickup geometry plays an important role in distortion.
@article{novak2020guitar,
    author={Novak, Antonin and Lihoreau, Bertrand and Brasseur, Emmanuel and Lotton, Pierrick and Simon, Laurent},
    title={Guitar Pickup: Measurement of its Two-Dimensional Nonlinear Function},
    journal={J. Audio Eng. Soc},
    volume={68},
    number={3},
    doi={10.17743/jaes.2020.0002},
    url={https://doi.org/10.17743/jaes.2020.0002},
    pages={184-192},
    year={2020},
    publisher={}
}

A. Novak & P. Cisar (2020), "Crayfish heart rate monitoring with an accelerometer", Ecol. Indic.. Vol. 111(0), pp. 2-6.

url
Crayfish are widely used as a bioindicator of water pollution levels and as a model organism to study environmental changes. Existing non-invasive systems for water pollution detection using the crayfish use the optical principle of heart rate monitoring. The aim of this study is to develop an alternative method to crayfish heart rate monitoring which solves the problems associated with the current methods. The introduced solution is based on the use of crayfish vibration measurement using an accelerometer. The method enables us to determine the heart rate and provide a quantitively similar signal to the existing systems. The advantage of this method is the direct detection of crayfish locomotor activity from a signal measured by one sensor only; optical sensors require an additional sensor for crayfish motion detection. The method is an alternative to the existing methods for water pollution detection and ethological experiments.
@article{Novak2020,
    author={Novak, Antonin and Cisar, Petr},
    title={Crayfish heart rate monitoring with an accelerometer},
    journal={Ecol. Indic.},
    volume={111},
    number={0},
    doi={10.1016/j.ecolind.2019.105993},
    url={https://doi.org/10.1016/j.ecolind.2019.105993},
    pages={2-6},
    year={2020},
    publisher={}
}

2019

A. Novak, P. Cisar, M. Bruneau, P. Lotton & L. Simon (2019), "Localization of sound-producing fish in a water-filled tank", J. Acoust. Soc. Am.. Vol. 146(6), pp. 4842-4850.

blog post PDF url
In this paper, the authors introduce an algorithm for locating sound-producing fish in a small rectangular tank that can be used, e.g., in behavioral bioacoustical studies to determine which fish in a group is sound-producing. The technique consists of locating a single sound source in the tank using signals gathered by four hydrophones placed in the tank together with a group of fish under study. The localization algorithm used in this paper is based on a ratio of two spectra ratios: the spectra ratio between the sound pressure measured by hydrophones at two locations and the spectra ratio between the theoretical Green’s functions at the same locations. The results are compared to a localization based on image processing technique and with video recordings acquired synchronously with the acoustic recordings.
@article{Novak2019fish,
    author={Novak, Antonin and Cisar, Petr and Bruneau, Michel and Lotton, Pierrick and Simon, Laurent},
    title={Localization of sound-producing fish in a water-filled tank},
    journal={J. Acoust. Soc. Am.},
    volume={146},
    number={6},
    doi={10.1121/1.5138607},
    url={https://doi.org/10.1121/1.5138607},
    pages={4842-4850},
    year={2019},
    publisher={Acoustical Society of America}
}

2018

A. Novak, M. Bruneau & P. Lotton (2018), "Small-Sized Rectangular Liquid-Filled Acoustical Tank Excitation: A Modal Approach Including Leakage Through the Walls", Acta Acustica united with Acustica. Vol. 104(4), pp. 586-596.

blog post PDF url
Small-sized water-filled tanks with thin (compared to the acoustic wavelength) walls that are surrounded on all sides by air, have been used over the past decades for underwater studies. So far, the analytical modelling of the acoustic field inside such an enclosure considers either the eigenvalue problem with pressure release conditions at the walls or imposes empirical impedance boundary conditions, or even proceeds in analogy with the field within a waveguide. At low frequencies, approximations involving an incompressible fluid or the Laplace equation have been used. Those models have limitations that are always caused by simplification of the boundary conditions. This paper deals with both an analytical formulation expressing the acoustic leakage through the walls (lossy and reacting walls) and modal solutions for the acoustic pressure and acoustic velocity fields when a source emits energy and when it is shut off, providing the transient acoustic response of small tanks quantitatively. Several analytical results are compared with experimental observations.
@article{novak2018small,
    author={Novak, Antonin and Bruneau, Michel and Lotton, Pierrick},
    title={Small-Sized Rectangular Liquid-Filled Acoustical Tank Excitation: A Modal Approach Including Leakage Through the Walls},
    journal={Acta Acustica united with Acustica},
    volume={104},
    number={4},
    doi={10.3813/AAA.919199},
    url={https://www.ingentaconnect.com/contentone/dav/aaua/2018/00000104/00000004/art00005?crawler=true&mimetype=application/pdf},
    pages={586-596},
    year={2018},
    publisher={S. Hirzel Verlag}
}

A. Novak, L. Simon & P. Lotton (2018), "A simple predistortion technique for suppression of nonlinear effects in periodic signals generated by nonlinear transducers", Journal of Sound and Vibration. Vol. 420(0), pp. 104-113.

blog post url
Mechanical transducers, such as shakers, loudspeakers and compression drivers that are used as excitation devices to excite acoustical or mechanical nonlinear systems under test are imperfect. Due to their nonlinear behaviour, unwanted contributions appear at their output besides the wanted part of the signal. Since these devices are used to study nonlinear systems, it should be required to measure properly the systems under test by overcoming the influence of the nonlinear excitation device. In this paper, a simple method that corrects distorted output signal of the excitation device by means of predistortion of its input signal is presented. A periodic signal is applied to the input of the excitation device and, from analysing the output signal of the device, the input signal is modified in such a way that the undesirable spectral components in the output of the excitation device are cancelled out after few iterations of real-time processing. The experimental results provided on an electrodynamic shaker show that the spectral purity of the generated acceleration output approaches 100 dB after few iterations (1 s). This output signal, applied to the system under test, is thus cleaned from the undesirable components produced by the excitation device; this is an important condition to ensure a correct measurement of the nonlinear system under test.
@article{novak2018simple,
    author={Novak, A and Simon, L and Lotton, P},
    title={A simple predistortion technique for suppression of nonlinear effects in periodic signals generated by nonlinear transducers},
    journal={Journal of Sound and Vibration},
    volume={420},
    number={0},
    doi={10.1016/j.jsv.2017.11.046},
    url={https://doi.org/10.1016/j.jsv.2018.01.038},
    pages={104-113},
    year={2018},
    publisher={Elsevier}
}

B. Maillou, P. Lotton, A. Novak & L. Simon (2018), "Modelling nonlinear viscoelastic behaviours of loudspeaker suspensions-like structures", Journal of Sound and Vibration. Vol. 416(0), pp. 213-223.

blog post url
Mechanical properties of an electrodynamic loudspeaker are mainly determined by its suspensions (surround and spider) that behave nonlinearly and typically exhibit frequency dependent viscoelastic properties such as creep effect. The paper aims at characterizing the mechanical behaviour of electrodynamic loudspeaker suspensions at low frequencies using nonlinear identification techniques developed in recent years. A Generalized Hammerstein based model can take into account both frequency dependency and nonlinear properties. As shown in the paper, the model generalizes existing nonlinear or viscoelastic models commonly used for loudspeaker modelling. It is further experimentally shown that a possible input-dependent law may play a key role in suspension characterization.
@article{maillou2018modelling,
    author={Maillou, Balbine and Lotton, Pierrick and Novak, Antonin and Simon, Laurent},
    title={Modelling nonlinear viscoelastic behaviours of loudspeaker suspensions-like structures},
    journal={Journal of Sound and Vibration},
    volume={416},
    number={0},
    doi={10.1016/j.jsv.2017.11.046},
    url={https://doi.org/10.1016/j.jsv.2017.11.046},
    pages={213-223},
    year={2018},
    publisher={Elsevier}
}

2017

A. Novak et al. (2017), "Measurements and modeling of the nonlinear behavior of a guitar pickup at low frequencies", Applied Sciences. Vol. 7(1), pp. 50.

PDF url
Description of the physical behavior of electric guitars is still not very widespread in the scientific literature. In particular, the physical models describing a nonlinear behavior of pickups still requires some refinements. The study presented in this paper is focused on nonlinear modeling of the pickups. Two main issues are raised. First, is the currently most used nonlinear model (a Hammerstein model) sufficient for the complex nonlinear behavior of the pickup? In other words, would a more complex model, such as a Generalized Hammerstein that can deal better with the nonlinear memory, yield better results? The second troublesome issue is how to measure the nonlinear behavior of a pickup correctly. A specific experimental set-up allowing for driving the pickup in a controlled way (string displacement perpendicular to the pickup) and to separate the nonlinear model of the pickup from other nonlinearities in the measurement chain is proposed. Thanks to this experimental set-up, a Generalized Hammerstein model of the pickup is estimated for frequency range 15–500 Hz and the results are compared with a simple Hammerstein model. A comparison with experimental results shows that both models succeed in describing the pickup when used in realistic conditions.
@article{novak2017measurements,
    author={Novak, Antonin and Guadagnin, Leo and Lihoreau, Bertrand and Lotton, Pierrick and Brasseur, Emmanuel and Simon, Laurent},
    title={Measurements and modeling of the nonlinear behavior of a guitar pickup at low frequencies},
    journal={Applied Sciences},
    volume={7},
    number={1},
    doi={10.3390/app7010050},
    url={https://www.mdpi.com/2076-3417/7/1/50},
    pages={50},
    year={2017},
    publisher={Multidisciplinary Digital Publishing Institute}
}

Y. Baccouche, M. Bentahar, C. Mechri, A. Novak & R. E. Guerjouma (2017), "Nonlinear analysis of damaged metal-based composite plates using guided waves", Acta Acustica united with Acustica. Vol. 103(6), pp. 967-977.

url
Resonance experiments have already proved the high level of nonlinearity in complex materials, including micro-cracked composites. However, the nonlinear parameters related to elastic modulus and damping are defined around a given resonance mode and are therefore known over a very limited frequency domain. In this contribution, the nonlinear parameters are determined on intact and damaged metal-based composite plates for several flexural resonances. Furthermore, the use of the theoretical formalism corresponding to guided flexural waves allowed to define a nonlinear parameter over a larger frequency domain. Finally, the nonlinear convolution method allowed to take advantage of the harmonics related to the different resonance modes in order to define new non-linear parameters whose sensitivity is much greater than the ones determined at the fundamental frequencies.
@article{baccouche2017nonlinear,
    author={Baccouche, Y and Bentahar, M and Mechri, C and Novak, A and Guerjouma, R El},
    title={Nonlinear analysis of damaged metal-based composite plates using guided waves},
    journal={Acta Acustica united with Acustica},
    volume={103},
    number={6},
    doi={https://doi.org/10.3813/AAA.919126},
    url={https://www.ingentaconnect.com/content/dav/aaua/2017/00000103/00000006/art00010},
    pages={967-977},
    year={2017},
    publisher={S. Hirzel Verlag}
}

M. Zakerin et al. (2017), "Thermal Characterization of Dynamic Silicon Cantilever Array Sensors by Digital Holographic Microscopy", Sensors. Vol. 17(6), pp. 11.

PDF url
In this paper, we apply a digital holographic microscope (DHM) in conjunction with stroboscopic acquisition synchronization. Here, the temperature-dependent decrease of the first resonance frequency (S1(T)) and Young’s elastic modulus (E1(T)) of silicon micromechanical cantilever sensors (MCSs) are measured. To perform these measurements, the MCSs are uniformly heated from T0 = 298 K to T = 450 K while being externally actuated with a piezo-actuator in a certain frequency range close to their first resonance frequencies. At each temperature, the DHM records the time-sequence of the 3D topographies for the given frequency range. Such holographic data allow for the extracting of the out-of-plane vibrations at any relevant area of the MCSs. Next, the Bode and Nyquist diagrams are used to determine the resonant frequencies with a precision of 0.1 Hz. Our results show that the decrease of resonance frequency is a direct consequence of the reduction of the silicon elastic modulus upon heating. The measured temperature dependence of the Young’s modulus is in very good accordance with the previously-reported values, validating the reliability and applicability of this method for micromechanical sensing applications.
@article{s17061191,
    author={Zakerin, Marjan and Novak, Antonin and Toda, Masaya and Emery, Yves and Natalio, Filipe and Butt, Hans-Jürgen and Berger, Rüdiger},
    title={Thermal Characterization of Dynamic Silicon Cantilever Array Sensors by Digital Holographic Microscopy},
    journal={Sensors},
    volume={17},
    number={6},
    doi={10.3390/s17061191},
    url={https://doi.org/10.3390/s17061191},
    pages={11},
    year={2017},
    publisher={}
}

A. Novak, P. Honzik & M. Bruneau (2017), "Dynamic behaviour of a planar micro-beam loaded by a fluid-gap: Analytical and numerical approach in a high frequency range, benchmark solutions", Journal of Sound and Vibration. Vol. 401(2017), pp. 36-53.

url
Miniaturized vibrating MEMS devices, active (receivers or emitters) or passive devices, and their use for either new applications (hearing, meta-materials, consumer devices, ...) or metrological purposes under non-standard conditions, are involved today in several acoustic domains. More in-depth characterisation than the classical ones available until now are needed. In this context, the paper presents analytical and numerical approaches for describing the behaviour of three kinds of planar micro-beams of rectangular shape (suspended rigid or clamped elastic planar beam) loaded by a backing cavity or a fluid-gap, surrounded by very thin slits, and excited by an incident acoustic field. The analytical approach accounts for the coupling between the vibrating structure and the acoustic field in the backing cavity, the thermal and viscous diffusion processes in the boundary layers in the slits and the cavity, the modal behaviour for the vibrating structure, and the non-uniformity of the acoustic field in the backing cavity which is modelled in using an integral formulation with a suitable Green's function. Benchmark solutions are proposed in terms of beam motion (from which the sensitivity, input impedance, and pressure transfer function can be calculated). A numerical implementation (FEM) is handled against which the analytical results are tested.
@article{novak2017dynamic,
    author={Novak, A and Honzik, P and Bruneau, M},
    title={Dynamic behaviour of a planar micro-beam loaded by a fluid-gap: Analytical and numerical approach in a high frequency range, benchmark solutions},
    journal={Journal of Sound and Vibration},
    volume={401},
    number={2017},
    doi={10.1016/j.jsv.2017.04.026},
    url={https://doi.org/10.1016/j.jsv.2017.04.026},
    pages={36-53},
    year={2017},
    publisher={Elsevier}
}

2016

A. Novak, F. Rund & P. Honzik (2016), "Impulse Response Measurements using MLS Technique on Nonsynchronous Devices", Journal of the Audio Engineering Society. Vol. 64(12), pp. 978-987.

url
Maximum-length sequences (MLS) are widely used for measurement of impulse responses of linear time-invariant systems in acoustic and audio systems. It is usually believed that one of the drawbacks of the MLS technique is a requirement of synchronized devices used for the generation and acquisition of the MLS signals. This study shows that the MLS technique can easily be improved and applied to devices that are not synchronous, or operate at different sampling frequencies. To show the efficiency of the proposed modification to MLS technique, the authors provided several experiments, as for example, a measurement with devices working at 44.1 kHz on the generation side and 96 kHz on the acquisition side, or a measurement of the frequency response function of an inexpensive mobile phone in which the synchronous clocking is not possible. Modifications to the classical MLS method are easy to implement and do not require excessive computational cost.
@article{novak2016impulse,
    author={Novak, Antonin and Rund, Frantisek and Honzik, Petr},
    title={Impulse Response Measurements using MLS Technique on Nonsynchronous Devices},
    journal={Journal of the Audio Engineering Society},
    volume={64},
    number={12},
    doi={10.17743/jaes.2016.0050},
    url={https://doi.org/10.17743/jaes.2016.0050},
    pages={978-987},
    year={2016},
    publisher={Audio Engineering Society}
}

A. Novak (2016), "Modeling Viscoelastic Properties of Loudspeaker Suspensions Using Fractional Derivatives", J. Audio Eng. Soc. Vol. 64(1), pp. 35-44.

url
This paper shows that mechanical models of the materials used in loudspeaker suspensions may be more accurate by incorporating fractional derivatives in the model. In conventional differential equations, the differentiating order is a real integer, whereas for fractional derivatives the order is only constrained to be a real number with a fractional piece. The results of four loudspeakers with different types of suspension show that the proposed model with a single fractional element provides very low RMS error when compared to the measured data for all loudspeakers measured in standard atmosphere as well as in vacuum. Many materials used in loudspeaker suspensions exhibit significant frequency dependence of damping and compliance due to their various viscoelastic properties. Many physical processes, including the viscoelastic materials, exhibit fractional order behavior. In addition there exists a physical interpretation of the fractional derivatives, which makes them more compelling that a purely empirical model.
@article{novak2016fractional,
    author={Novak, Antonin},
    title={Modeling Viscoelastic Properties of Loudspeaker Suspensions Using Fractional Derivatives},
    journal={J. Audio Eng. Soc},
    volume={64},
    number={1},
    doi={10.17743/jaes.2015.0091},
    url={https://doi.org/10.17743/jaes.2015.0091},
    pages={35-44},
    year={2016},
    publisher={}
}

2015

A. Novak, P. Lotton & L. Simon (2015), "Synchronized Swept-Sine: Theory, Application, and Implementation", Journal of the Audio Engineering Society. Vol. 63(10), pp. 786-798.

url
Exponential swept-sine signals are very often used to analyze nonlinear audio systems. A reexamination of this methodology shows that a synchronization procedure is necessary for the proper analysis of higher harmonics. An analytical expression of spectra of the swept-sine signal is derived and used in the deconvolution of the impulse response. Matlab code for generation of the synchronized swept-sine, deconvolution, and separation of the impulse responses is given. This report provides a discussion of some application issues and an illustrative example of harmonic analysis of current distortion of a woofer. An analysis of the higher harmonics of the current distortion of a woofer is compared using both the synchronized and the non-synchronized swept-sine signals.
@article{novak2015synchronized,
    author={Novak, Antonin and Lotton, Pierrick and Simon, Laurent},
    title={Synchronized Swept-Sine: Theory, Application, and Implementation},
    journal={Journal of the Audio Engineering Society},
    volume={63},
    number={10},
    doi={10.17743/jaes.2015.0071},
    url={https://doi.org/10.17743/jaes.2015.0071},
    pages={786-798},
    year={2015},
    publisher={Audio Engineering Society}
}

B. Merit & A. Novak (2015), "Magnet-Only Loudspeaker Magnetic Circuits: A Solution for Significantly Lower Current Distortion", Journal of the Audio Engineering Society. Vol. 63(6), pp. 463-474.

url
When the motor of a classical electrodynamic loudspeaker has been replaced by a so-called magnet-only motor, current distortion caused by magnetic circuits is significantly reduced. All of the nonlinear soft-iron pieces are removed, and the magnetic field is created by a combination of permanent magnets. Two loudspeakers are compared in the study: the first sample is a mass-produced 6.5-inch loudspeaker. The second sample is built replacing the magnetic circuit of the first loudspeaker by the magnet-only structure. Thus, the moving parts, including the voice-coil, diaphragm, and suspensions are kept identical for both samples, leading to a comparison based only on their motors. The measurements on the original loudspeaker have shown a significant variation of apparent resistance and inductance of the voice-coil with current, displacement, and frequency. Conversely, the measurements on the magnet-only loudspeaker have shown almost no variation whatever the frequency, current level, or position of the voice-coil.
@article{merit2015magnet,
    author={Merit, Benoit and Novak, Antonin},
    title={Magnet-Only Loudspeaker Magnetic Circuits: A Solution for Significantly Lower Current Distortion},
    journal={Journal of the Audio Engineering Society},
    volume={63},
    number={6},
    doi={10.17743/jaes.2015.0051},
    url={https://doi.org/10.17743/jaes.2015.0051},
    pages={463-474},
    year={2015},
    publisher={Audio Engineering Society}
}

2014

A. Novak, B. Maillou, P. Lotton & L. Simon (2014), "Nonparametric Identification of Nonlinear Systems in Series", Instrumentation and Measurement, IEEE Transactions on. Vol. 63(8), pp. 2044-2051.

PDF url
In this paper, a method allowing the identification of two nonlinear systems in series is presented. More precisely, the identification of the second nonlinear subsystem under test is achieved by considering the effects of the nonlinearities of the first subsystem. The method is based on the estimation of the higher harmonic frequency responses from the measurement of distorted input and output signals. The second nonlinear system is then modeled by nonparametric generalized Hammerstein model made up of power series associated with linear filters. The method is experimentally validated in the well-known framework of nonlinear propagation of acoustic waves.
@article{novak2014nonparametric,
    author={Novak, Antonin and Maillou, Balbine and Lotton, Pierrick and Simon, Laurent},
    title={Nonparametric Identification of Nonlinear Systems in Series},
    journal={Instrumentation and Measurement, IEEE Transactions on},
    volume={63},
    number={8},
    doi={10.1109/TIM.2014.2303552},
    url={https://doi.org/10.1109/TIM.2014.2303552},
    pages={2044-2051},
    year={2014},
    publisher={IEEE}
}

2012

J. B. Legland, V. Tournat, O. Dazel, A. Novak & V. Gusev (2012), "Linear and nonlinear Biot waves in a noncohesive granular medium slab: Transfer function, self-action, second harmonic generation", The Journal of the Acoustical Society of America. Vol. 131(6), pp. 4292-4303.

PDF url
Experimental results are reported on second harmonic generation and self-action in a noncohesive granular medium supporting wave energy propagation both in the solid frame and in the saturating fluid. The acoustic transfer function of the probed granular slab can be separated into two main frequency regions: a low frequency region where the wave propagation is controlled by the solid skeleton elastic properties, and a higher frequency region where the behavior is dominantly due to the air saturating the beads. Experimental results agree well with a recently developed nonlinear Biot wavemodel applied to granular media. The linear transfer function, second harmonic generation, and self-action effect are studied as a function of bead diameter, compaction step, excitation amplitude, and frequency. This parametric study allows one to isolate different propagation regimes involving a range of described and interpreted linear and nonlinear processes that are encountered in granular media experiments. In particular, a theoretical interpretation is proposed for the observed strong self-action effect.
@article{legland2012linear,
    author={Legland, Jean, Baptiste and Tournat, Vincent and Dazel, Olivier and Novak, Antonin and Gusev, Vitalyi},
    title={Linear and nonlinear Biot waves in a noncohesive granular medium slab: Transfer function, self-action, second harmonic generation},
    journal={The Journal of the Acoustical Society of America},
    volume={131},
    number={6},
    doi={10.1121/1.4712020},
    url={https://doi.org/10.1121/1.4712020},
    pages={4292-4303},
    year={2012},
    publisher={Acoustical Society of America}
}

A. Novak, M. Bentahar, V. Tournat, R. El Guerjouma & L. Simon (2012), "Nonlinear acoustic characterization of micro-damaged materials through higher harmonic resonance analysis", NDT & E International. Vol. 45(1), pp. 1-8.

PDF url
A method developed for the analysis of nonlinear systems is applied for the first time to non-destructive testing of diverse materials using vibrations and elastic waves. This method allows to extract the vibrational/acoustical responses of the system, at the excitation frequency and importantly, also at higher harmonics, with the help of a nonlinear convolution signal analysis. It is then possible to make use of the robust nonlinear resonance method together with the harmonic generation method in order to analyze the nonlinear elastic resonances of a sample at excitation frequency harmonics. Definitions of the nonlinear hysteretic parameters associated to higher harmonic resonances are provided. The bases of the signal analysis method are also described. A higher sensitivity to the presence of gradual damage compared to the classical nonlinear resonance method is demonstrated experimentally for diverse materials and configurations.
@article{novak2012nonlinear,
    author={Novak, A and Bentahar, M and Tournat, V and El Guerjouma, R and Simon, L},
    title={Nonlinear acoustic characterization of micro-damaged materials through higher harmonic resonance analysis},
    journal={NDT & E International},
    volume={45},
    number={1},
    doi={10.1016/j.ndteint.2011.09.006},
    url={https://doi.org/10.1016/j.ndteint.2011.09.006},
    pages={1-8},
    year={2012},
    publisher={Elsevier}
}

2010

A. Novak, M. Bentahar, L. Simon & R. El Guerjouma (2010), "Apport de la convolution non lineaire a la caracterisation de l'endommagement des materiaux heterogenes en resonance", Instrumentation Mesures Metrologie. Vol. 10(1), pp. 175-194.

PDF url
Implementation of nonlinear convolution method, usually called Synchronous Swept Sine, is presented in the frame of damage characterisation of heterogeneous materials. The estimation of higher order frequency responses, the used approach allows detection of new resonance modes created during the interaction of classical resonance modes with damage. The created modes made possible the definition of new nonlinear parameters whose sensitivity to the presence of damage as well as its propagation is likely to go beyond structural health monitoring applications by offering a better understanding of the physics corresponding to hysteretic non linear materials.
@article{novak2010_I2M,
    author={Novak, A. and Bentahar, M. and Simon, L. and El Guerjouma, R.},
    title={Apport de la convolution non lineaire a la caracterisation de l'endommagement des materiaux heterogenes en resonance},
    journal={Instrumentation Mesures Metrologie},
    volume={10},
    number={1},
    doi={10.3166/i2m.10.1-2.175-194},
    url={https://doi.org/10.3166/i2m.10.1-2.175-194},
    pages={175-194},
    year={2010},
    publisher={}
}

A. Novak, L. Simon & P. Lotton (2010), "Analysis, synthesis, and classification of nonlinear systems using synchronized swept-sine method for audio effects", EURASIP Journal on Advances in Signal Processing. Vol. 2010(0), pp. 4.

PDF url
A new method of identification, based on an input synchronized exponential swept-sine signal, is used to analyze and synthesize nonlinear audio systems like overdrive pedals for guitar. Two different pedals are studied; the first one exhibiting a strong influence of the input signal level on its input/output law and the second one exhibiting a weak influence of this input signal level. The Synchronized Swept Sine method leads to a Generalized Polynomial Hammerstein model equivalent to the pedals under test. The behaviors of both pedals are illustrated through model-based resynthesized signals. Moreover, it is also shown that this method leads to a criterion allowing the classification of the nonlinear systems under test, according to the influence of the input signal levels on their input/output law.
@article{novak2010analysis,
    author={Novak, Antonin and Simon, Laurent and Lotton, Pierrick},
    title={Analysis, synthesis, and classification of nonlinear systems using synchronized swept-sine method for audio effects},
    journal={EURASIP Journal on Advances in Signal Processing},
    volume={2010},
    number={0},
    doi={10.1155/2010/793816},
    url={https://doi.org/10.1155/2010/793816},
    pages={4},
    year={2010},
    publisher={Hindawi Publishing Corp.}
}

A. Novak, L. Simon, F. Kadlec & P. Lotton (2010), "Nonlinear system identification using exponential swept-sine signal", Instrumentation and Measurement, IEEE Transactions on. Vol. 59(8), pp. 2220-2229.

PDF url
In this paper, we propose a method for nonlinear system (NLS) identification using a swept-sine input signal and based on nonlinear convolution. The method uses a nonlinear model, namely, the nonparametric generalized polynomial Hammerstein model made of power series associated with linear filters. Simulation results show that the method identifies the nonlinear model of the system under test and estimates the linear filters of the unknown NLS. The method has also been tested on a real-world system: an audio limiter. Once the nonlinear model of the limiter is identified, a test signal can be regenerated to compare the outputs of both the real-world system and its nonlinear model. The results show good agreement between both model-based and real-world system outputs.
@article{novak2010nonlinear,
    author={Novak, Antonin and Simon, Laurent and Kadlec, Frantisek and Lotton, Pierrick},
    title={Nonlinear system identification using exponential swept-sine signal},
    journal={Instrumentation and Measurement, IEEE Transactions on},
    volume={59},
    number={8},
    doi={10.1109/TIM.2009.2031836},
    url={https://doi.org/10.1109/TIM.2009.2031836},
    pages={2220-2229},
    year={2010},
    publisher={IEEE}
}



International conferences, workshops

2023

A. Novak & P. Honzik (2023), "Measurement of Microphone Harmonic Distortion using Predistortion Technique", Proc. Forum Acusticum, Turin, Italy

PDF
Measurement of microphone nonlinear behavior has always been a challenging task due to a lack of linear acoustic sources. A recently developed technique that corrects a distorted output signal of an excitation device using predistortion of its input signal is used in this work. Thanks to this technique, a spectrally clean sinusoidal acoustic pressure is generated inside a closed box in which the microphone under test is placed together with a low-distortion reference microphone. The goal of this paper is twofold. First, show that the direct measurement of distortion is possible even for high-pressure levels; second, compare the harmonic distortion of several types of microphones such as studio electrostatic and electrodynamic microphones, MEMS microphones and laboratory microphones.
@inproceedings{novak2023coil,
    author={Novak, Antonin and Honzik, Petr},
    title={Measurement of Microphone Harmonic Distortion using Predistortion Technique},
    booktitle={Proc. Forum Acusticum},
    address={Turin, Italy},
    year={2023},
    organization={EAA}
}

A. Novak & P. Lotton (2023), "Loudspeaker voice coil in various magnetic environments - a practical approach for students", Proc. Forum Acusticum, Turin, Italy

PDF
After teaching transducer measurements to master's students for several years, we noticed that the voice coil inductance and the associated physical phenomena inside the loudspeaker motor are frequently considered the most challenging concepts for students to grasp. Consequently, we developed a 3-hour practical lab session to address this topic, where students are guided to measure the voice coil in various configurations. The voice coil impedance is measured in various conditions, including air, in the disassembled speaker motor (without magnet), inside conductive but non-magnetic materials, with another coil shorted in its proximity, and inside the assembled motor. The objective of the lab is to gain a clear understanding of the influence of each phenomenon related to permeability, currents, and other related factors. The purpose of this document is twofold: first, to serve as a support material for students; and second, to share our university laboratory experience with other institutions.
@inproceedings{novak2023coil,
    author={Novak, Antonin and Lotton, Pierrick},
    title={Loudspeaker voice coil in various magnetic environments - a practical approach for students},
    booktitle={Proc. Forum Acusticum},
    address={Turin, Italy},
    year={2023},
    organization={EAA}
}

2022

O. Munroe, S. Letourneur & A. Novak (2022), "Design of an electronic circuit for loudspeaker real-time digital signal processing", 16eme Congrès Français d'Acoustique, Marseille, France

PDF url
In modern audio systems, real-time digital signal processing algorithms are widely used for a variety of applications. The possibility of using a simple electronic circuit for variety of research projects has shown remarkable potential and is gradually attracting more and more attention from researchers and engineers. This contribution describes a design of such a board used in the framework of a PhD thesis whose subject is centred on the real-time correction of loudspeaker nonlinearities. The solution chosen in this work is based on a Teensy 3.6 microcontroller which is easy to program using the Arduino IDE and the libraries provided by Teensy. Two solutions are provided: one with an Audio board available on the market and another with a homemade board. Both solutions contain two inputs and at least one output (all 16 bits). This contribution does not detail the compensation algorithm related to the loudspeaker nonlinearities but focuses on the boards design, comparison of proposed solutions, and provides the basic codes to perform the real-time digital signal processing.
@inproceedings{novakCFA2022,
    author={Munroe, O. and Letourneur, S. and Novak, A.},
    title={Design of an electronic circuit for loudspeaker real-time digital signal processing},
    booktitle={16eme Congrès Français d'Acoustique},
    address={Marseille, France},
    year={2022},
    organization={SFA}
}

2020

A. Novak (2020), "Compression and expansion nonlinear effects in an electrodynamic loudspeaker: experiments vs. model failure ", Proc. Forum Acusticum, Lyon, France

PDF
An electrodynamic loudspeaker is known to be a nonlinear device due to many physical phenomena resulting in position and current dependent loudspeaker parameters. The most common position-dependent nonlinear sources are the force factor Bl(x) and stiffness K_ms(x). These nonlinear effects have many consequences, the most commonly used in measurements being harmonic distortion and intermodulation distortion. Another consequence is the compression effect of the dynamics. Due to a lower Bl(x) and a higher K_ms(x) for larger displacements, one would expect a compression behavior, i. e. a decreasing ratio of X/U or X/I with increasing excitation level (X, I, and U are the displacement, current, and voltage, respectively). Nevertheless, measurements in the low frequencies of many loudspeakers show an expansion behavior, i.e., an increase in the X/U and X/I ratio with increasing excitation level. This non-intuitive behavior, which runs counter to the fundamental theory of simple models of Bl(x) and K_ms(x), is studied and discussed in this paper.
@inproceedings{novak2020compression,
    author={Novak, Antonin},
    title={Compression and expansion nonlinear effects in an electrodynamic loudspeaker: experiments vs. model failure },
    booktitle={Proc. Forum Acusticum},
    address={Lyon, France},
    year={2020},
    organization={EAA}
}

S. Belloncle, T. Julienne & A. Novak (2020), "Loudspeaker characterization by indirect measurement of diaphragm velocity", Audio Engineering Society Convention 149, AES Show 2020

url
A complete characterization of an electrodynamic transducers consists, among others, of measurement of the membrane velocity. Modern measurement devices, such as laser vibrometers and accelerometers, are coming with their set of disadvantages. The laser devices can be costly, the accelerometers, being attached to the device under test, are influencing the measurement and are unusable for small-sized loudspeakers. This paper explores the indirect measurement of the membrane velocity, using a single microphone, and the limits for measurement of loudspeaker parameters, including the frequency-dependent ones due to creep effects and eddy currents.
@inproceedings{belloncle2020loudspeaker,
    author={Belloncle, Simon and Julienne, Thibaut and Novak, Antonin},
    title={Loudspeaker characterization by indirect measurement of diaphragm velocity},
    booktitle={Audio Engineering Society Convention 149},
    address={AES Show 2020},
    year={2020},
    organization={}
}

A. Ramirez, V. Tokala, A. Novak, F. Ablitzer & M. Melon (2020), "Bistable Digital Audio Effect", Proc. of the 23th Int. Conference on Digital Audio Effects (eDAFx-20/21), Vienna, Austria

PDF url
A mechanical system is said to be bistable when its moving parts can rest at two equilibrium positions. The aim of this work is to model the vibration behaviour of a bistable system and use it to create a sound effect, taking advantage of the nonlinearities that characterize such systems. The velocity signal of the bistable system excited by an audio signal is the output of the digital effect. The latter is coded in C++ language and compiled into VST3 format that can be run as an audio plugin within most of the commercial digital audio workstation software in the market and as a standalone application. A Web Audio API demonstration is also available online as a support material.
@inproceedings{ramirez10bistable,
    author={Ramirez, Alexander and Tokala, Vikas and Novak, Antonin and Ablitzer, Frederic and Melon, Manuel},
    title={Bistable Digital Audio Effect},
    booktitle={Proc. of the 23th Int. Conference on Digital Audio Effects (eDAFx-20/21)},
    address={Vienna, Austria},
    year={2020},
    organization={}
}

2019

A. Novak, P. Cisar, M. Bruneau, P. Lotton & L. Simon (2019), "Finding a sound-producing fish in a water-fi lled tank with a few hydrophones", XXVII International Bioacoustics Congress (IBAC), Brighton, England

url
In this contribution, we present a recently developed algorithm for localization of sound-producing fish in a small rectangular tank. The algorithm is based on a theoretical acoustic model of the fish tank and can be used for behavioral bioacoustical studies to determine which fish in a group is sound-producing. We use four hydrophones placed in the tank together with a group of fish under study (6 sparkling gouramis). To show the accuracy of the localization, we compare the results with a localization based on image processing technique and with video recordings acquired synchronously with the acoustic recordings.
@inproceedings{novak2019ibac,
    author={Novak, Antonin and Cisar, Petr and Bruneau, Michel and Lotton, Pierrick and Simon, Laurent},
    title={Finding a sound-producing fish in a water-fi lled tank with a few hydrophones},
    booktitle={XXVII International Bioacoustics Congress (IBAC)},
    address={Brighton, England},
    year={2019},
    organization={}
}

A. Novak (2019), "Measurement of loudspeaker parameters, A pedagogical approach", 23rd International Congress on Acoustics, Aachen, Germany

blog post PDF
@inproceedings{ica19,
    author={Novak, Antonin},
    title={Measurement of loudspeaker parameters, A pedagogical approach},
    booktitle={23rd International Congress on Acoustics},
    address={Aachen, Germany},
    year={2019},
    organization={}
}

A. Novak, L. Simon, P. Lotton & M. Melon (2019), "Predistortion Technique for Generating Spectrally Clean Excitation Signals for Audio and Electro-Acoustic Nonlinear Measurements", Audio Engineering Society Convention 146, Dublin, Ireland

blog post url
In many audio and electro-acoustic nonlinear measurements we need to excite the nonlinear system under test with an excitation device that is not linear. A typical example is the study of the nonlinear behavior of a loudspeaker mechanical part, where the mechanical part (the nonlinear system under test) is excited externally, either with a shaker or pneumatically using another loudspeaker. We often consider that the excitation device is linear, which is unfortunately not correct. In this paper we present a simple method that corrects the distorted output signal of the excitation device by pre-distorting the input signal. The process is based on harmonic injection and can be applied to any periodic signal that is used for the measurement, e.g., a sine wave to measure the total harmonic distortion (THD), a two-tone signal to measure an intermodulation distortion (IMD), or a multi-tone signal. The experimental results provided on an electrodynamic loudspeaker show that the undesired spectral components of the acoustic pressure inside the sealed box can be suppressed to the level of the background noise.
@inproceedings{novak2019predistortion,
    author={Novak, Antonin and Simon, Laurent and Lotton, Pierrick and Melon, Manuel},
    title={Predistortion Technique for Generating Spectrally Clean Excitation Signals for Audio and Electro-Acoustic Nonlinear Measurements},
    booktitle={Audio Engineering Society Convention 146},
    address={Dublin, Ireland},
    year={2019},
    organization={Audio Engineering Society}
}

J. Cadavid & A. Novak (2019), "Practical Problems in Building Parametric Loudspeakers with Ultrasonic Piezoelectric Emitters", Audio Engineering Society Convention 146, Dublin, Ireland

url
In this paper we deal with some practical issues that one can encounter when building a parametric loudspeaker with ultrasonic piezoelectric emitters. We measured several of those transducers (with resonance frequency 40 kHz) available on the market, observing a strong nonlinear behavior of many of them. We also tested a hundred of piezoelectric emitters of the same series and studied the influence of the standard deviation of the resonance frequency and the sensitivity on the performance of the parametric loudspeaker. We conclude that, when constructing a parametric loudspeaker with low-cost piezoelectric emitters, the individual behavior of each of them should be considered. This allows to minimize the effect of their differences and, thus, improve the quality of the sound generated.
@inproceedings{cadavid2019practical,
    author={Cadavid, Jose and Novak, Antonin},
    title={Practical Problems in Building Parametric Loudspeakers with Ultrasonic Piezoelectric Emitters},
    booktitle={Audio Engineering Society Convention 146},
    address={Dublin, Ireland},
    year={2019},
    organization={Audio Engineering Society}
}

2018

B. Kakonyi, R. Abdul Jaleel & A. Novak (2018), "Balloon explosion, wood-plank, revolver shot, or traditional loudspeaker large-band excitation: which is better for microphone measurement?", Audio Engineering Society Convention 144, Milan, Italy

url
The work presented in this paper investigates different sound sources used for microphone measurement in an anechoic room. The microphone under test is measured using three different physical impulse-like sources: balloon explosions, sounds created using wood-planks, and revolver shots; and the results are compared with a measurement in which the sound is generated by a loudspeaker excited with a large band swept-sine signal. The frequency response functions of three different commercial microphones are measured and compared with the data provided by the manufacturer.
@inproceedings{kakonyi2018balloon,
    author={Kakonyi, Balazs and Abdul Jaleel, Riyas and Novak, Antonin},
    title={Balloon explosion, wood-plank, revolver shot, or traditional loudspeaker large-band excitation: which is better for microphone measurement?},
    booktitle={Audio Engineering Society Convention 144},
    address={Milan, Italy},
    year={2018},
    organization={Audio Engineering Society}
}

A. Novak, B. Lihoreau, P. Lotton, E. Brasseur & L. Simon (2018), "Experimental Study Of Guitar Pickup Nonlinearity", Proc. of the 18th Int. Conference on Digital Audio Effects (DAFx-18), Aveiro, Portugal

PDF url
In this paper, we focus on studying the nonlinear behavior of the pickup of an electric guitar and on its modeling. The approach is purely experimental, based on physical assumptions and attempts to find a nonlinear model that, with few parameters, would be able to predict the nonlinear behavior of the pickup. In our experimental setup, a piece of string is attached to a shaker and vibrates perpendicularly to the pickup in the frequency range between 60 Hz and 400 Hz. The oscillations are controlled by a linearization feedback to create a purely sinusoidal steady state movement of the string. In the first step, harmonic distortions of three different magnetic pickups (a single-coil, a humbucker, and a rail-pickup) are compared to check if they provide different distortions. In the second step, a static nonlinearity of Paiva's model is estimated from experimental signals. In the last step, the pickup nonlinearities are compared and an empirical model that fits well all three pickups is proposed.
@inproceedings{novak_dafx18,
    author={Novak, Antonin and Lihoreau, Bertrand and Lotton, Pierrick and Brasseur, Emmanuel and Simon, Laurent},
    title={Experimental Study Of Guitar Pickup Nonlinearity},
    booktitle={Proc. of the 18th Int. Conference on Digital Audio Effects (DAFx-18)},
    address={Aveiro, Portugal},
    year={2018},
    organization={}
}

2016

B. Maillou, P. Lotton, A. Novak & L. Simon (2016), "Comportement Non-Linéaire des Suspensions Mécaniques d'un Haut-Parleur : Caractérisation et Identification", 13e Congrès Français d'Acoustique, Le Mans, France

Dans le fonctionnement d'un haut-parleur électrodynamique, les non-linéarités générées en cas de fort niveau d'excitation électrique peuvent être d'origines électrique, mécanique ou acoustique. Le travail présenté ici porte sur le comportement mécanique en basses fréquences de l'équipage mobile lorsqu'il est soumis à un déplacement de grande amplitude, et plus spécifiquement sur le comportement des suspensions. Traditionnellement, le modèle de Thiele & Small décrit le comportement mécanique de l'équipage mobile en régime linéaire comme celui d'un système masse-ressort amorti. En régime non-linéaire, les paramètres de raideur et d'amortissement de ce modèle ne suffisent plus pour décrire le comportement mécanique complexe observé. L'identification d'un modèle « Hammerstein Généralisé », à partir de données expérimentales caractérisant un équipage mobile d'un haut-parleur découplé de la partie "moteur magnétique", est alors proposée. Les performances et limites d'un tel modèle non-linéaire non-paramétrique des suspensions de haut-parleur sont discutées. Enfin, les paramètres d’un modèle dit de « Thiele et Small étendu » sont déduits à partir des paramètres du modèle «Hammerstein Généralisé» et permettent de mettre en évidence l’évolution de la raideur et de l’amortissement avec la fréquence d’excitation et le déplacement de la membrane, ainsi que la dépendance des phénomènes observés avec le niveau d’excitation.
@inproceedings{novakCFA16_04,
    author={Maillou, B. and Lotton, P. and Novak, A. and Simon, L.},
    title={Comportement Non-Linéaire des Suspensions Mécaniques d'un Haut-Parleur : Caractérisation et Identification},
    booktitle={13e Congrès Français d'Acoustique},
    address={Le Mans, France},
    year={2016},
    organization={}
}

A. Novak, P. Honzik, M. Bruneau & F. Fohr (2016), "Étude de filtre pour protecteur auditif passif", 13e Congrès Français d'Acoustique, Le Mans, France

Les protecteurs auditifs passifs placés dans le canal auditif sont d'usage courant pour assurer la protection contre les bruits intenses. De nos jours, les différents principes utilisés permettent généralement de réaliser des atténuations globales de 15 dB à 25 dB mais avec une courbe de restitution en fréquence très déséquilibrée (atténuation principalement des hautes fréquences et suppression de la résonance naturelle du canal auditif vers 3 kHz). Seuls les principes de filtres acoustiques utilisant une fine membrane fermant le canal présentent une réponse en fréquence plus équilibrée. Cependant, ce principe présente l'inconvénient majeur d’empêcher la ventilation du canal auditif (transpiration) et l'égalisation de pression. Pour tenter de proposer une autre voie de développement de filtres acoustiques à réponse en fréquence adaptée, nous proposons une étude qui a pour objet d'associer des petits éléments dont le comportement relève des fonctions de petits guides et cavités de formes adaptées, couplés entre eux, formant ainsi un système qui ne nécessite pas l'usage de membrane (tout au plus celui d'un petit oscillateur). Les résultats analytiques et numériques obtenus à ce jour, qui prennent en compte la dissipation thermo-visqueuse dans les couches limites, montrent la faisabilité de tels protecteurs auditifs.
@inproceedings{novakCFA16_03,
    author={Novak, A. and Honzik, P. and Bruneau, M. and Fohr, F.},
    title={Étude de filtre pour protecteur auditif passif},
    booktitle={13e Congrès Français d'Acoustique},
    address={Le Mans, France},
    year={2016},
    organization={}
}

A. Novak, P. Honzik, M. Bruneau & N. Joly (2016), "Modélisation analytique et contrôle numérique du comportement d'une micro-poutre vibrante chargée par une couche de fluide à l'arrière", 13e Congrès Français d'Acoustique, Le Mans, France

L'usage de systèmes dynamiques miniatures (MEMS), actifs (récepteurs ou émetteurs) ou passifs, pour de nouvelles applications (audition, méta-matériaux, biens de consommations, ...) ou en métrologie sous des conditions non-standard, apparaît à l'heure actuelle dans divers domaines de l'acoustique. Des caractérisations plus précises que celles disponibles aujourd'hui sont rendues nécessaires. Dans ce contexte, le travail présenté ici porte sur une approche analytique et numérique pour décrire le comportement des diverses formes de systèmes comportant chacun une micro-poutre de forme rectangulaire (poutre rigide suspendue ou poutre élastique plane) chargée à l'arrière par une petite cavité ou une fine couche de fluide, entourée de très fines fentes (qui évitent la conception complexe de suspensions, assurent l'équilibre des pressions statiques et contribuent à l'amortissement), et mise en vibration par un champ acoustique incident. La modélisation analytique prend en compte le fort couplage entre la poutre et le champ de pression acoustique à l'arrière, les processus de diffusion visqueux et thermiques dans les couches limites dans les fentes et la cavité arrière, le comportement modal dans le cas de la poutre élastique, la non uniformité du champ acoustique à l'arrière de la poutre qui est modélisée en faisant usage du formalisme intégral, et finalement le couplage des champs acoustiques de part et d'autre de la poutre au travers des fentes. Dans les diverses configurations étudiées, des solutions générales sont proposées pour le mouvement de la poutre, conduisant à exprimer une sensibilité, une impédance d'entrée ou une fonction de transfert acoustique selon le système considéré en pratique. Un calcul numérique (FEM) est mené en parallèle, qui permet de valider les résultats analytiques obtenus.
@inproceedings{novakCFA16_02,
    author={Novak, A. and Honzik, P. and Bruneau, M. and Joly, N.},
    title={Modélisation analytique et contrôle numérique du comportement d'une micro-poutre vibrante chargée par une couche de fluide à l'arrière},
    booktitle={13e Congrès Français d'Acoustique},
    address={Le Mans, France},
    year={2016},
    organization={}
}

E. Brasseur, A. Novak, P. Lotton, B. Lihoreau & a. L. (2016), "Développement de bancs de mesure pour la caractérisation des pickups électromagnétiques", 13e Congrès Français d'Acoustique, Le Mans, France

Les microphones (pickups) de guitares électriques sont des transducteurs électromagnétiques sensibles à la vitesse des cordes de la guitare. Dans le cadre des recherches menées au LAUM concernant la caractérisation de ces capteurs, un premier banc de mesure a été développé. Ce banc permet de mesurer simultanément la tension de sortie du pickup et la vitesse d’une corde, lorsque celle-ci est soumise à une excitation impulsionnelle effectuée à l'aide d'un plectre automatisé. Un deuxième dispositif expérimental a été développé permettant de mesurer la réponse du pickup lorsqu’une portion de corde, montée sur un pot vibrant et placée devant le capteur, est soumise à un mouvement sinusoïdal entretenu contrôlé en fréquence et en amplitude. Le comportement du pot-vibrant étant non-linéaire, la vitesse de la corde est distordue, ce qui introduit des erreurs non négligeables dans l’analyse des résultats expérimentaux. Par conséquent un système de contrôle actif a été développé qui consiste à ajouter, en temps réel, à la tension d'excitation du pot vibrant des harmoniques supérieurs, contrôlés en amplitude et en phase de manière à générer une vitesse de la corde purement sinusoïdale. Afin de maîtriser au mieux les paramètres des pickups à caractériser, l’étude est menée sur une série de pickups monoplots réalisés au laboratoire. Les différentes étapes de conception des bancs de mesures, les contraintes techniques, les choix technologiques ainsi que le processus de réalisation des pickups seront présentés. La présentation du système de contrôle actif permettant la linéarisation de la vitesse générée par la source du deuxième banc de mesure fera l'objet d'un point particulier.
@inproceedings{novakCFA16_01,
    author={Brasseur, E. and Novak, A. and Lotton, P. and Lihoreau, B. and Guadagnin L.},
    title={Développement de bancs de mesure pour la caractérisation des pickups électromagnétiques},
    booktitle={13e Congrès Français d'Acoustique},
    address={Le Mans, France},
    year={2016},
    organization={}
}

A. Novak, L. Simon & P. Lotton (2016), "Extension of Generalized Hammerstein Model to Non-Polynomial Inputs", EUSIPCO 2016, Budapest, Hungary

PDF
The Generalized Hammerstein model has been successfully used during last few years in many physical applications to describe the behavior of a nonlinear system under test. The main advantage of such a nonlinear model is its capability to model efficiently nonlinear systems while keeping the computational cost low. On the other hand, this model can not predict complicated nonlinear behaviors such as hysteretic one. In this paper, we propose an extension of the Generalized Hammerstein model to a model with non polynomial nonlinear inputs that allows modeling more complicated nonlinear systems. A simulation provided in this paper shows a good agreement between the model and the hysteretic nonlinear system under test.
@inproceedings{novak2016extension,
    author={Novak, Antonin and Simon, Laurent and Lotton, Pierrick},
    title={Extension of Generalized Hammerstein Model to Non-Polynomial Inputs},
    booktitle={EUSIPCO 2016},
    address={Budapest, Hungary},
    year={2016},
    organization={}
}

A. Novak et al. (2016), "Non-linear Identification Of An Electric Guitar Pickup", Proc. of the 16th Int. Conference on Digital Audio Effects (DAFx-16), Brno, Czech Republic

PDF
Physical models of electric guitars are still not very widespread in the scientific literature. Especially, the description of the non linear behavior of pickups still requires some refinements. This paper deals with the identification of pickup non linearities based on a Hammerstein representation, by means of a specific experimental set-up to drive the pickup in a controlled way. A comparison with experimental results shows that the model succeeds in describing the pickup when used in realistic conditions.
@inproceedings{novak_dafx16,
    author={Novak, A. and Guadagnin, L. and Lihoreau, B. and Lotton, P. and Brasseur, E. and Simon, L.},
    title={Non-linear Identification Of An Electric Guitar Pickup},
    booktitle={Proc. of the 16th Int. Conference on Digital Audio Effects (DAFx-16)},
    address={Brno, Czech Republic},
    year={2016},
    organization={}
}

2014

A. Novak, P. Lotton & L. Simon (2014), "Comportement Non Linéaire des Suspensions d'un Haut-parleur: Mesure Dynamique à l'Aide d'une Méthode de Contrôle Actif", 12e Congrès Français d'Acoustique, Poitiers, France

Une nouvelle technique non destructive pour mesurer les paramètres de suspension (raideur Kms et résistance mécanique Rms) d'un haut-parleur en régime non linéaire à l'aide d'une méthode de contrôle actif est présentée. Le but du contrôle actif est d'éliminer les harmoniques supérieures contenues dans le signal de déplacement de telle sorte qu'un mouvement purement harmonique de l'equipage mobile soit assuré. La raideur non linéaire Kms est alors mesurée en fonction du déplacement instantanée et du niveau d'excitation; la résistance mécanique Rms est mesurée en fonction de la vitesse. Les résultats de mesure d'un haut-parleur électrodynamique présentés dans ce papier montrent une dépendance importante de la résistance mécanique Rms avec la vitesse de membrane. De plus, les résultats montrent que la raideur Kms ne dépend pas seulement du déplacement instantané comme généralement pris en compte dans la modélisation de haut-parleur non linéaire, mais aussi du niveau d'excitation ou autrement dit du déplacement crête. La raideur Kms et la résistance mécanique Rms sont caractérisées à plusieurs fréquences, afin d'étudier la dépendence fréquencielle de ces paramétres.
@inproceedings{novakCFA14,
    author={Novak, A. and Lotton, P. and Simon, L.},
    title={Comportement Non Linéaire des Suspensions d'un Haut-parleur: Mesure Dynamique à l'Aide d'une Méthode de Contrôle Actif},
    booktitle={12e Congrès Français d'Acoustique},
    address={Poitiers, France},
    year={2014},
    organization={}
}

A. Novak, P. Lotton & L. Simon (2014), "Nonlinear Force Factor Measurement of an Electrodynamic Loudspeaker", Proc. Forum Acusticum, Krakow, Poland

PDF
An electrodynamic loudspeaker is usually characterized using a linear lumped parameter model, whose physical parameters are assumed to be constant. Such a linear approximation is sufficient for small excitation levels, but becomes insufficient for larger ones. The model is therefore usually extended allowing a variation of the parameters with the displacement of the diaphragm, the current intensity in the coil, the frequency and many other physical quantities. One of the important nonlinear parameters is the force factor Bl that characterizes the magneto-electrical performance of the loudspeaker and that can be highly dependent on the voice-coil displacement. In this paper, we propose a simple and precise method to estimate the force factor Bl of the loudspeaker as a function of the displacement of the voice-coil. The voice-coil is displaced from its rest position electrically using a direct current and a swept-sine signal of small level is fed to the loudspeaker. The signals of the electric tension, the current and the velocity of the diaphragm are recorded and used to estimated the force factor Bl. This process is repeated with different values of direct current to obtain the force factor as a static function of the voice-coil displacement. The experimental results of arbitrary chosen loudspeakers are shown.
@inproceedings{novak2014force,
    author={Novak, Antonin and Lotton, Pierrick and Simon, Laurent},
    title={Nonlinear Force Factor Measurement of an Electrodynamic Loudspeaker},
    booktitle={Proc. Forum Acusticum},
    address={Krakow, Poland},
    year={2014},
    organization={EAA}
}

A. Novak, P. Lotton & L. Simon (2014), "Dynamic measurement of loudspeaker suspension parameters using an active harmonic control technique", Audio Engineering Society Convention 136, Berlin, Germany

url
A new nondestructive technique to measure the nonlinear suspension parameters (stiffness Kms and mechanical resistance Rms) of a loudspeaker using an active harmonic control technique is presented. The goal of the active harmonic control is to eliminate the higher harmonics from the displacement signal so that a purely harmonic motion of the diaphragm is ensured. The nonlinear stiffness Kms is then measured as a function of instantaneous and peak displacement; the mechanical resistance Rms is measured as a function of velocity. A frequency dependence of these parameters is also discussed.
@inproceedings{novak2014dynamic,
    author={Novak, Antonin and Lotton, Pierrick and Simon, Laurent},
    title={Dynamic measurement of loudspeaker suspension parameters using an active harmonic control technique},
    booktitle={Audio Engineering Society Convention 136},
    address={Berlin, Germany},
    year={2014},
    organization={Audio Engineering Society}
}

B. Maillou, P. Lotton, A. Novak & L. Simon (2014), "Mechanical Nonlinearities of Electrodynamic Loudspeakers: An Experimental Study", Audio Engineering Society Convention 136, Berlin, Germany

url
Spider and surround suspensions are at the origin of viscoelastic and nonlinear behaviors of loudspeakers because of their assembly geometry and their intrinsic materials. We propose here a new dynamic experimental method to characterize these properties. We drive the loudspeaker moving part with a shaker and measure the driving force, the acceleration, the velocity, and the displacement. Results are presented and discussed for a given loudspeaker, which surround suspensions exhibit viscoelastic behavior.
@inproceedings{maillou2014mechanical,
    author={Maillou, Balbine and Lotton, Pierrick and Novak, Antonin and Simon, Laurent},
    title={Mechanical Nonlinearities of Electrodynamic Loudspeakers: An Experimental Study},
    booktitle={Audio Engineering Society Convention 136},
    address={Berlin, Germany},
    year={2014},
    organization={Audio Engineering Society}
}

2012

A. Novak & B. Merit (2012), "Static and Dynamic Measurement of a Magnet-only Loudspeaker", EURONOISE 2012, Prague, Czech Republic

PDF
A new loudspeaker technology using Magnet-only motors has been presented recently. Unlike the classical electrodynamic loudspeaker, no iron is used in the motor part. The presence of iron in traditional loudspeaker motor is a source of nonlinearities and, as a consequence, deteriorates the reproduced sound quality. According to the theory of Magnet-only motors, the loudspeaker parameters should vary less than in the case of a classical electrodynamic loudspeaker. In this paper, the parameters of a Magnet-only loudspeaker are measured statically and dynamically. The results presented in this paper clarify why a magnet-only loudspeaker improves the quality of the reproduced sound.
@inproceedings{novak2012euronoise,
    author={Novak, Antonin and Merit, Benoit},
    title={Static and Dynamic Measurement of a Magnet-only Loudspeaker},
    booktitle={EURONOISE 2012},
    address={Prague, Czech Republic},
    year={2012},
    organization={EURONOISE}
}

A. Novak & B. Merit (2012), "Magnet-only loudspeaker motors: linear behavior theory vs. Nonlinear measurements", Acoustics 2012 Nantes, Nantes, France

PDF url
A few years ago, a new concept of magnet-only loudspeaker has been proposed to improve the quality of the reproduced sound. Such a loudspeaker is called magnet-only because its magnetic circuit is totally made of rare-earth permanent magnets. Unlike the classical electrodynamic loudspeaker, no iron is used. According to the theory, the exclusive use of permanent magnets and the absence of iron can lead to a uniform motor parameters (force factor Bl, resistance Re and inductance Le) over the voice-coil displacement and thus to a decrease of the nonlinear distortion and to an improvement of the sound quality. To our knowledge, such motor parameters have not been consistently quantified, neither their variations. In this paper, the variation of the parameters of a magnet-only loudspeaker are measured. The goal is to verify the theory and to show that using a simple measurement procedure one can understand better why a magnet-only loudspeaker improves the quality of the reproduced sound.
@inproceedings{novak2012magnet,
    author={Novak, Antonin and Merit, Benoit},
    title={Magnet-only loudspeaker motors: linear behavior theory vs. Nonlinear measurements},
    booktitle={Acoustics 2012 Nantes},
    address={Nantes, France},
    year={2012},
    organization={}
}

B. Maillou et al. (2012), "Electrodynamic loudspeakers suspensions nonlinearities, study and measurements", Acoustics 2012 Nantes, Nantes, France

url
Nonlinearities characterization and identification of acoustical systems is an active research field, notably in the area of electroacoustics. An innovative characterization and identification method, based on the nonlinear convolution and the use of a synchronized swept-sine excitation signal, has recently been developped. In this paper, this method is applied to the study of the nonlinearities exhibited by the mechanical suspensions of an electrodynamic loudspeaker. Nonlinear behaviours are specially studied as function of the amplitude of the displacement of the membrane of the loudspeaker. In this context, we develop an experimental setup from the signal generator to the mobile part of the tested component, specially designed in order to obtain a great harmonic distortion measurement dynamic range and, thus, to allow the measurement of weak nonlinearities. Nonlinearities are estimated for both the external and internal suspensions independently, and for several kinds of loudspeakers. Results are compared with those obtained by other methods available in the literature.
@inproceedings{maillou2012electrodynamic,
    author={Maillou, Balbine and Novak, Antonin and Blondeau, James and Brouard, Bruno and Génevaux, Jean-Michel and Lotton, Pierrick and Simon, Laurent},
    title={Electrodynamic loudspeakers suspensions nonlinearities, study and measurements},
    booktitle={Acoustics 2012 Nantes},
    address={Nantes, France},
    year={2012},
    organization={}
}

2011

P. Michaud, P. Herzog, S. Meunier & A. Novak (2011), "Perceptual evaluation of loudspeaker nonlinearities", Proc. Forum Acusticum, Aalborg, Denmark

PDF url
This study deals with the evaluation of the perception of distortion in the restitution of musical excerpts played by loudspeakers. The main focus is to evaluate the perceived distortion induced by low frequency nonlinearities of the loudspeaker. A large panel is created in which only the distortion varies among the loudspeakers. The "virtual" loudspeakers are created from one single actual loudspeaker by associating its anechoic filtered tweeter recording and its woofer signal model in which the distortion can be modified. To model the woofer restitution, its identification is done using a swept sine technique which separates the linear and higher order impulse responses. These impulse responses are used as linear and nonlinear frequency filters characterizing the sound pressure of the system under test. The woofer synthesis consists in expanding the musical excerpts on the basis of Chebyshev polynomials combined with the identified frequency filters (Novak et al., DAFx 2010). Different "virtual" loudspeakers are obtained by amplifying or reducing the influence of specific nonlinearities. In order to evaluate the perceived distortion, in realistic listening conditions, the auralization technique was used to convolve the "virtual" loudspeakers with the impulse response of a suitable room. Upcoming listening tests are conducted with headphones to evaluate the perceived differences over the created panel of loudspeakers with a method allowing the evaluation of large sets of stimuli. The dissimilarity judgments can then be analyzed with a Multidimensional scaling (MDS) technique to reveal the underlying dimensions of the distortion space.
@inproceedings{michaud2011perceptual,
    author={Michaud, Pierre-Yohan and Herzog, Philippe and Meunier, Sabine and Novak, Antonin},
    title={Perceptual evaluation of loudspeaker nonlinearities},
    booktitle={Proc. Forum Acusticum},
    address={Aalborg, Denmark},
    year={2011},
    organization={}
}

2010

A. Novak, M. Bentahar, V. Tournat, R. El Guerjouma & L. Simon (2010), "Can nonlinear convolution improve damage characterization using acoustic methods ?", 5eme Colloque Interdisciplinaire en Instrumentation, Le Mans, France

PDF
This work presents an original damage characterisation method of homogeneous and heterogeneous solids using a signal processing-based experimental approach. Convolution nonlinear method is used in order to improve experimental observations allowing to measure simultaneously the well known vibration spectrum, classically found using analysers, and harmonic spectrums which are out of reach when using the same analysers. The experimental approach has been validated to characterise a progressive damage corresponding to materials of different shape and nature (glass beads, polymer-based composite plates).
@inproceedings{C2I,
    author={Novak, A. and Bentahar, M. and Tournat, V. and El Guerjouma, R. and Simon, L.},
    title={Can nonlinear convolution improve damage characterization using acoustic methods ?},
    booktitle={5eme Colloque Interdisciplinaire en Instrumentation},
    address={Le Mans, France},
    year={2010},
    organization={}
}

A. Novak, L. Simon, P. Lotton & J. Gilbert (2010), "Nonlinear Analysis and Modeling of Electrodynamic Loudspeakers Based on Exponential Swept Sine Input Signal", 10th French Congress of Acoustics, Lyon, France

PDF url
Many real-world systems exhibit nonlinear behavior that must be taken into account when modeling such systems. In practise, especially in the field of electro-acoustics, nonlinearities appear with increasing input level. These nonlinearities are usually modeled by adding nonlinear parameters to the linear model, but the determination of nonlinear physical parameters is rather difficult problem. The authors present a simple method for analysis and identification of nonlinear systems, based on swept-sine signal excitation. The method is based on nonlinear convolution, firstly proposed by Angelo Farina [1], which serves for analysis of nonlinear amplitude characteristics. The result of the nonlinear convolution method is the set of harmonic distortion products (higher order nonlinear impulse responses) that can be expressed either in the form of the separated impulse responses or in the form of frequency dependent components. The original method [1] was improved in order to analyze also the phase charceristics of all the nonlinear parts. Therefore, the excitation sweep signal from the improved method [2] has to be setup very properly. The precisely measured nonlinear amplitude and phase characteristics can be used for analysis of the nonlinear system under test, or for identification of a nonlinear model (i.e. generalized Hammerstein model). The method is presented here when studying an electrodynamic loudspeaker from the nonlinear point of view. More precisely the electrodynamic loudspeaker under test is characterized by its electrical impedance leading to Thiele/Small parameters in a nonlinear framework. [1] A. Farina, ''Simultaneous measurement of impulse response and distortion with a swept-sine technique'', in Proc. 108th AES Convention, Paris, February, 2000. [2] A. Novak et al., ''Nonlinear System Identification Using Exponential Swept Sine Signal'', accepted for publication in IEEE Instrumentation and Measurement, 2009
@inproceedings{CFA10_01,
    author={Novak, A. and Simon, L. and Lotton, P. and Gilbert, J.},
    title={Nonlinear Analysis and Modeling of Electrodynamic Loudspeakers Based on Exponential Swept Sine Input Signal},
    booktitle={10th French Congress of Acoustics},
    address={Lyon, France},
    year={2010},
    organization={}
}

A. Novak, M. Bentahar, V. Tournat, R. El Guerjouma & L. Simon (2010), "Strutural Health Monitoring Based on Nonlinear System Identification", 10th French Congress of Acoustics, Lyon, France

In classical nonlinear resonance experiments, we generally use of a gain/phase analyzer to excite and display the materials Frequency Response Function FRF. However, the latter response is obtained using a tracking filter around the fundamental excitation frequency. In that case, the obtained information is generally a combination of the desired response and a contribution of a power series of harmonics responses. Moreover, the filtering procedure prevents from having higher harmonics responses whose knowledge is important in nonlinear damage characterization. In this contribution we propose a nonlinear system identification method based on an exponential swept-sine drive signal. The recorded response of the material under test is convolved with an inverse filter to obtain the nonlinear responses corresponding to the harmonics usually eliminated by the gain/phase analyzer filtering system. Therefore, a single swept-sine experiment allows to estimate higher harmonics responses (amplitude and phase) depending on the nonlinear behaviour of the material. The first measurements performed on a damaged polymer-based composite reveal that higher harmonics, generally neglected in nonlinear damage characterization because of the tracking filter, are very sensitive to damage and hence deserve more attention than usual. As the sensors stand by the structure this method could be used advantageously in Structural Health Monitoring.
@inproceedings{CFA10_02,
    author={Novak, A. and Bentahar, M. and Tournat, V. and El Guerjouma, R. and Simon, L.},
    title={Strutural Health Monitoring Based on Nonlinear System Identification},
    booktitle={10th French Congress of Acoustics},
    address={Lyon, France},
    year={2010},
    organization={}
}

F. Kadlec, K. Ulovec & A. Novak (2010), "Application of SWT and MISO Methods for Measurement and Analysis of Electro-Acoustic and Audio Systems", 17th Internation congress on Sound and Vibration, Cairo, Egypt

PDF url
This contribution deals with performance evaluation of electro-acoustic and audio systems with non-linearities by a combination of two existing methods: Sweep Sine Wave Signal Technique (SWT) and Multiple Input Single Output (MISO) nonlinear modeling. The signal, driving a system under test, is a logarithmically swept sine wave. Transfer functions of the system under test may be obtained from the response of the system and the swept sine wave "inverse filter" response. These transfer functions are further processed. Then the system under test can be considered as a MISO nonlinear model, consisting of a parallel combination of nonlinear branches, containing filters and memory-less power-law distortion functions. Such a model facilitates determination of a nonlinear response to the input signal. This method has been successfully used for analysis of various components of electro-acoustic and audio systems. In the first application described in our contribution, this method was used for a detailed analysis of transfer functions of a complete electro-acoustic system including an artificial head. This system is used for the analysis of spatial perception as well as determination of a signal compression impact on its perception. In the other application, this test method was used for measurement and evaluation of the whole transmission chain, from studio all the way to the listener. The signal source was a studio recorded CD that was transmitted by FM radio to the listener's receiver, where it was demodulated, monitored and analyzed. A good correlation with traditional methods of measurement was found.
@inproceedings{icsv17,
    author={Kadlec, F. and Ulovec, K. and Novak, A.},
    title={Application of SWT and MISO Methods for Measurement and Analysis of Electro-Acoustic and Audio Systems},
    booktitle={17th Internation congress on Sound and Vibration},
    address={Cairo, Egypt},
    year={2010},
    organization={}
}

A. Novak, L. Simon, P. and Lotton & J. Gilbert (2010), "Chebyshev Model and Synchronized Swept Sine Method in Nonlinear Audio Effect Modeling", Proc. of the 13th Int. Conference on Digital Audio Effects (DAFx-10), Graz, Austria

PDF url
A new method for the identification of nonlinear systems, based on an input exponential swept sine signal has been proposed by Farina ten years ago. This method has been recently modified in purpose of nonlinear model estimation using a synchronized swept sine signal. It allows a robust and fast one-path analysis and identification of the unknown nonlinear system under test. In this paper this modified method is applied with Chebyshev polynomial decomposition. The combination of the Synchronized Swept Sine Method and Chebyshev polynomials leads to a nonlinear model consisting of several parallel branches, each branch containing a nonlinear Chebyshev polynomial following by a linear filter. The method is tested on an overdrive effect pedal to simulate an analog nonlinear effect in digital domain.
@inproceedings{novak_dafx10,
    author={Novak, A. and Simon, L. and and Lotton, P and Gilbert, J.},
    title={Chebyshev Model and Synchronized Swept Sine Method in Nonlinear Audio Effect Modeling},
    booktitle={Proc. of the 13th Int. Conference on Digital Audio Effects (DAFx-10)},
    address={Graz, Austria},
    year={2010},
    organization={}
}

2009

Novak A., Bentahar B., Tournat V., Simon L. & (2009), "Sensitive Damage Characterization Based on Nonlinear System Identification", ICNEW XIV, Lisbon, Portugal

@inproceedings{ICNEW,
    author={Novak A., and Bentahar B., and Tournat V., and Simon L., and El Guerjouma R.},
    title={Sensitive Damage Characterization Based on Nonlinear System Identification},
    booktitle={ICNEW XIV},
    address={Lisbon, Portugal},
    year={2009},
    organization={}
}

A. Novak, L. Simon, F. Kadlec & P. Lotton (2009), "Modeling of Nonlinear Audio Systems using Swept-Sine signals: Application to audio effects", Proc. of the 12th Int. Conference on Digital Audio Effects (DAFx-09), Como, Italy

PDF
In this paper a new method for analysis and modeling of nonlinear audio systems is presented. The method is based on swept-sine excitation signal and nonlinear convolution firstly presented in [1, 2]. It can be used in nonlinear processing for audio applications, to simulate analog nonlinear effects (distortion effects, limiters) in digital domain.
@inproceedings{novak2009modeling,
    author={Novak, A. and Simon, L. and Kadlec, F. and Lotton, P},
    title={Modeling of Nonlinear Audio Systems using Swept-Sine signals: Application to audio effects},
    booktitle={Proc. of the 12th Int. Conference on Digital Audio Effects (DAFx-09)},
    address={Como, Italy},
    year={2009},
    organization={}
}

2008

A. Novak, F. Kadlec, L. Simon & P. Lotton (2008), "Direct Path MISO method: Identification of Nonlinear Electro-acoustic Systems", Proceedings of Workshop 2008, Prague, Czech Republic

PDF
@inproceedings{worshop08,
    author={Novak, A. and Kadlec, F. and Simon, L. and Lotton, P.},
    title={Direct Path MISO method: Identification of Nonlinear Electro-acoustic Systems},
    booktitle={Proceedings of Workshop 2008},
    address={Prague, Czech Republic},
    year={2008},
    organization={Czech Technical University in Prague}
}

A. Novak, L. Simon, F. Kadlec & P. Lotton (2008), "A New Method for Identification of Nonlinear Systems Using MISO Model with Swept-Sine Technique: Application to Loudspeaker Analysis", 124th AES Convention, Amsterdam, Netherlands

url
This work presents a Multiple Input Single Output (MISO) nonlinear model in combination with sine-sweep signals as a method for nonlinear system identification. The method is used for identification of loudspeaker nonlinearities and can be applied to nonlinearities of any audio components. It extends the method based on nonlinear convolution presented by Farina, providing a nonlinear model that allows to simulate the identified nonlinear system. The MISO model consists of a parallel combination of nonlinear branches containing linear filters and memory-less power-law distortion functions. Once the harmonic distortion components are identified by the method of Farina, the linear filters of the MISO model can be derived. The practical application of the method is demonstrated on a loudspeaker.
@inproceedings{aes08,
    author={Novak, A. and Simon, L. and Kadlec, F. and Lotton, P.},
    title={A New Method for Identification of Nonlinear Systems Using MISO Model with Swept-Sine Technique: Application to Loudspeaker Analysis},
    booktitle={124th AES Convention},
    address={Amsterdam, Netherlands},
    year={2008},
    organization={}
}

A. Novak, L. Simon, J. Gilbert, P. Lotton & F. Kadlec (2008), "Comparison of non Linear System Identification Methods : the Example of non Linear Propagation of Acoustic Waves in Ducts", Proceedings of Acoustics'08 Paris, Paris, France

url
The weakly non linear propagation of travelling acoustic waves in ducts is a well known problem leading to approximated analytical solutions. From an experimental point of view, the classical way for estimating the non linear parameters of propagation is to generate sine waves and to analyse the higher order harmonics as a function of the amplitude and the frequency of the excitation. In this work, new methods for estimating the non linear parameters of propagation are developed and compared to the sine excitation based method. The excitation signals associated to these new methods can be stationary noise or logarithmic chirps. For these excitation signals, the data processing is based on Multiple Input Single Output (MISO) direct path method. The comparison is made in terms of signal to noise ratio robustness and computation time. Experimental and theoretical results are also compared. We particularly show that a measurement using only one logarithmic chirp allows estimating accurate results for a broad band of amplitudes and frequencies. Associated computational time is then drastically reduced.
@inproceedings{Acoustics08,
    author={Novak, A. and Simon, L. and Gilbert, J. and Lotton, P. and Kadlec, F.},
    title={Comparison of non Linear System Identification Methods : the Example of non Linear Propagation of Acoustic Waves in Ducts},
    booktitle={Proceedings of Acoustics'08 Paris},
    address={Paris, France},
    year={2008},
    organization={}
}

F. Kadlec & A. Novak (2008), "Measurement and Analysis of Non-linear Distortion of Electro-acoustic and Audio Systems", 15th Internation congress on Sound and Vibration, Daejeon, Korea

PDF
A novel technique for analysis of nonlinear distortion of electro-acoustic and audio systems is presented. The method is based on the combination of a Sweep sine Wave signal Technique (SWT) and a Multiple Input Single Output (MISO) nonlinear modeling. The signal driving a system under test is a logarithmic sine wave sweep. The response of the system under test is acquired and used in a convolution with swept sine wave "inverse filter" response to obtain a set of transfer functions. This set of transfer functions provides information about harmonic distortion. These transfer functions are further processed. Then the system under test can be modeled as a MISO nonlinear model, consisting of a parallel combination of nonlinear branches containing filters and memory-less power-law distortion functions. Such a model facilitates determination of a nonlinear response to the input signal. This method was used to analyze transfer functions of various components of electro-acoustic systems. It was applied to testing of recording and reproduction equipment for audio signals, including equipment for their further digital signal processing. During testing of multichannel systems, the method enables evaluation of not only individual channels, but also their mutual interaction. A good correlation with traditional methods of distortion measurement has been shown.
@inproceedings{ICSV15,
    author={Kadlec, F. and Novak, A.},
    title={Measurement and Analysis of Non-linear Distortion of Electro-acoustic and Audio Systems},
    booktitle={15th Internation congress on Sound and Vibration},
    address={Daejeon, Korea},
    year={2008},
    organization={}
}

2007

A. Novak (2007), "Identification of Nonlinear Systems: Volterra Series", Poster 2007, Prague, Czech Republic

PDF
Traditional measurement of multimedia systems such as linear impulse response, transfer function, are sufficient but not faultless. For these methods the pure linear system is considered. Nonlinearities, which are usually included in the most of real systems are disregarded. One of the simple methods that can describe the nonlinear system used in practice is coefficient of distortion or intermodulation distortion, but these methods cannot be used to determine nonlinearities themselves. This paper describe one of the methods to identify nonlinear systems called Volterra Series. A simplification for this method is proposed and an experiment with audio amplifier is shown to test this method.
@inproceedings{poster03,
    author={Novak, A.},
    title={Identification of Nonlinear Systems: Volterra Series},
    booktitle={Poster 2007},
    address={Prague, Czech Republic},
    year={2007},
    organization={}
}

A. Novak (2007), "Identification of Nonlinear Systems: MISO modeling", Poster 2007, Prague, Czech Republic

PDF
Abstract. In this paper, the fundamentals of analysis of nonlinear systems using the direct path MISO method (Multiple Input / Single Output ) with the input signal as a record from stationary gaussian (ergodic) random process with zero mean value will be introduced. The MISO technique for linear system is well known in many fields of science and technical applications and the first publications using MISO technique for nonlinear systems was presented by J.S.Bendat. This technique has been used in some acoustical applications (for example in vibration analysis). This paper will discuss the application of MISO method for general nonlinear system using a polynomial series. An easy experiment will be conducted to show the functionality of this method.
@inproceedings{poster02,
    author={Novak, A.},
    title={Identification of Nonlinear Systems: MISO modeling},
    booktitle={Poster 2007},
    address={Prague, Czech Republic},
    year={2007},
    organization={}
}

A. Novak (2007), "Identification of Nonlinearity of Electro-acoustic Systems using a Direct Path MISO Method", 19th International Congress on Acoustics, Madrid, Spain

PDF
The work deals with the Direct Path Multiple-Input Single-Output (MISO) method adapted for the identification of Nonlinearity of Electro-acoustic Systems. The method is based on a blind identification which uses decorrelated power series expansion, without having any knowledge of the shape of the nonlinear function. The nonlinear Direct Path model is represented by an equivalent Multiple-Input Single-Output linear model, where the inputs are nonlinear contributions of the original input signal. Each branch of a complex nonlinear model, with its nonlinear input, represents a "separable nonlinearity" or a static nonlinearity that is followed by a linear system so that the memory effect is represented by a linear filter. The input signal for identification is a record from stationary Gaussian random process. The method has been numerically tested on static nonlinear systems, such as limiter and death-zone systems. Also, an experiment on real electro-acoustic system has been performed.
@inproceedings{ica01,
    author={Novak, A.},
    title={Identification of Nonlinearity of Electro-acoustic Systems using a Direct Path MISO Method},
    booktitle={19th International Congress on Acoustics},
    address={Madrid, Spain},
    year={2007},
    organization={}
}

F. Kadlec & A. Novak (2007), "Comparision of Measurements Method for Analysis Electro-acoustic Systems with Non-linearities", 19th International Congress on Acoustics, Madrid, Spain

PDF
The great level of excitation of electro-acoustic systems may cause a nonlinear distortion of signal, which has an effect on its perception. For analysing and monitoring functionality of such a system there are two methods, Multiple-Input Single-Output method for nonlinear systems and a method utilizing sweep sine wave signals, which are compared in this contribution. The first method uses a stationary Gaussian random process as the input signal. Then a nonlinear system is represented by a multiple inputs linear system and additional nonlinear inputs are determined according to decorrelated power series expansion of the main input signal. The second method uses a logarithmic sweep signal as the input signal. The impulse response of the system is obtained by a convolution of the system response to the input sweep signal and the "inverse filter". The impulse response consists of several distorted partial components. These two methods have been compared from the accuracy of the algorithmic point of view and also by measuring by an electro-acoustic system.
@inproceedings{ica02,
    author={Kadlec, F. and Novak, A.},
    title={Comparision of Measurements Method for Analysis Electro-acoustic Systems with Non-linearities},
    booktitle={19th International Congress on Acoustics},
    address={Madrid, Spain},
    year={2007},
    organization={}
}

A. Novak & F. Rund (2007), "Systematicke chyby pri analyze nelinearnich systemu", Technical Computing Prague 2007, Prague, Czech Republic

PDF
@inproceedings{humsoft2007novak,
    author={Novak, A. and Rund, F.},
    title={Systematicke chyby pri analyze nelinearnich systemu},
    booktitle={Technical Computing Prague 2007},
    address={Prague, Czech Republic},
    year={2007},
    organization={}
}

2006

A. Novak (2006), "Digital Loudspeaker Arrays Analysis", Poster 2006, Prague, Czech Republic

PDF
This paper deals with the properties of loudspeaker with direct digital to analog conversion, the basic goal being to replace D/A converter , amplifier , and classic speakers. The advantages of this system should be lower operating costs and greater applicability. The main part of this paper is aimed at analyzing the signals acquired from the theoretical model of the digital loudspeaker arrays (DLA) with regards to the arangement of elements on the transducer field and the point of observation.
@inproceedings{poster01,
    author={Novak, A.},
    title={Digital Loudspeaker Arrays Analysis},
    booktitle={Poster 2006},
    address={Prague, Czech Republic},
    year={2006},
    organization={}
}

F. Rund & A. Novak (2006), "Problem spatne synchronizace vzorkovacich kmitoctu u MLS signalu: Model v prostredi MATLAB", Technical Computing Prague 2006, Prague, Czech Republic

PDF
@inproceedings{humsoft06,
    author={Rund, F. and Novak, A.},
    title={Problem spatne synchronizace vzorkovacich kmitoctu u MLS signalu: Model v prostredi MATLAB},
    booktitle={Technical Computing Prague 2006},
    address={Prague, Czech Republic},
    year={2006},
    organization={}
}



Book Chapters

2015

M. Bentahar et al. (2015), "Nonlinear acoustics and acoustic emission methods to monitor damage in mesoscopic elastic materials", Emerging Technologies in Non-Destructive Testing VI, CRC Press

url
Microcracked composites and metals usually exhibit a high level of nonlinearity in their elastic response already at low amplitudes of excitation. In order to quantify these behaviors, different nonlinear indicators can be used depending on the required parameters and the experimental configuration of interest. In this contribution several acoustic nonlinear techniques are presented in order to detect and monitor the presence and evolution of micro-cracks in different kind of materials. In particular, the development of a particular "vibrational/acoustical" arrangement made the use of the robust nonlinear resonance method together with the harmonic generation method possible. Besides, the definition of a proper nonlinear indicator and its dependence on the excitation amplitude, allowed obtaining a power law behavior of cracked materials. Finally, the correlation between acoustic emission and nonlinear relaxation results shows that all mechanisms relax as the logarithm of their corresponding energy but with different dynamics.
@incollection{bentahar2015nonlinear,
    author={Bentahar, M and El Guerjouma, R and Mechri, C and Baccouche, Y and Idjimarene, S and Novak, A and Toumi, S and Tournat, V and Thomas, JH and Scalerandi, M},
    title={Nonlinear acoustics and acoustic emission methods to monitor damage in mesoscopic elastic materials},
    year={2015},
}



Ph.D. Thesis

A. Novak (2009), "Identification of Nonlinear Systems in Acoustics", Universite du Maine, Le Mans, France AND Czech Technical University, Prague, Czech Republic. Ph.D. thesis.

PDF
The theory of linear time-invariant (LTI) systems has been extensively studied over decades and the estimation of any unknown LTI system, knowing both the input and output of the system, is a solved problem. Nevertheless, almost all real-world devices exhibit more or less nonlinear behavior. In the case of very weak nonlinearities, a linear approximation can be used. If the nonlinearities are stronger, the linear approximation fails and systems have to be described using a nonlinear model. The goal of this thesis is to design and develop simple methods for nonlinear systems identification that would be accurate and robust enough to be applicable for analysis and identification of nonlinear systems in several domains, even if the main focus here is on the domain of audio and acoustics. The goal is to identify a nonlinear system and find its generic nonlinear model in such way that the response of the model to any input signal would be the same as the one of the real-world nonlinear system under test. Two methods are developed in the thesis. Both methods are based on Multiple Input – Single Output (MISO) model. The model consists of several parallel branches, each branch consisting of two separated blocks: a nonlinear static function and a linear dynamic filter. The first method uses a white Gaussian noise as the excitation signal for the identification. This method is successfully tested on several simulation examples, but fails when identifying real world nonlinear systems. The second method is based on the nonlinear convolution and uses swept sine excitation signal. This method is successfully tested on several simulation examples. Moreover, it is theoretically shown that it could be used for the identification of systems exhibiting specific dynamical hysteresis (called hysteresis with viscosity-type effect). Two well known real world nonlinear systems (an audio limiter and an acoustic waveguide) are used to validate the second method. The validation is based on the comparison between the output of these real world systems and the output of their estimated models, when excited with the same input signal. The comparison is performed both subjectively, using a simple visual comparison in time or frequency domains, and objectively, using a relative mean square error criterion. Once validated, the method is used in the general frame of the study of electrodynamic loudspeaker quality. Preliminary results show that this method could be used for the nonlinearities loudspeakers identification, and that an inverse filtering minimizing these nonlinearities could possibly be performed with the help of this method.
@phdthesis{,
    author={Novak, A.},
    title={Identification of Nonlinear Systems in Acoustics},
    school={Universite du Maine, Le Mans, France AND Czech Technical University, Prague, Czech Republic}
    year={2009},
}