Projects for students

Auditory distance perception plays a major role in spatial awareness, enabling location of objects and avoidance of obstacles in the environment. Normal-hearing people can easily determine the approximate distance between themselves and a sound source by using only auditory cues, even when visual information is not available. The primary cues are sound level, reverberation, and frequency. Following hearing loss, the use of auditory level as a distance cue remains robust, while the reverberation cue becomes less effective. Previous studies have not found evidence that hearing-aid processing affects perceived auditory distance. This project aims to evaluate the impact of hearing aids on the distance-perception ability of individuals with hearing impairment.
The project consists of two stages. The first stage involves building a fully-equipped listening test experimental facility. This includes a loudspeaker and an electric rail to move it with precision. While the loudspeaker will generate different stimuli at varying distances, the rail will carry it to predetermined positions automatically and reliably. The second stage will conduct a listening test on a group of approximately 10 individuals with normal hearing and 20 people with hearing impairment.

Profile: Electrical engineering, Physics, Mechanics

Prerequisites: Acoustics, Psychoacoustic (not necessary),

Learning outcome: auditory distance perception knowledge and listening test experience

Context: Experiment setup (40%), listening tests (60%)

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, SPH, Electrical Engineering,

Memberships:

Hervé Lissek – Role: Owner, Qin Liu – Role: Supervisor,

A wind instrument can generally be assimilated to a axisymmetric hollow duct, with varying (or constant) cross-section along the main axis. One extremity of the duct may be open or closed (the mouth, that can be a reed – see a saxophone -, a mouthpiece – see a trombone – or an open hole – see the flute), whereas the other end (the bell) is generally horn-shaped.
It is well known that the timbre of the instrument is associated with the shape of the bell, the type of flow excitation by the lips at the mouthpiece, and, more importantly, to the boundary conditions at the two extremities (open/open, closed/open): this specifies the nature of resonances inside the main duct, which yields a certain series of resonance frequencies (harmonics).
Then, it is noticeable that the wind families (brass, wood, etc.) are not referring to the material in which it is made, but rather on the way sound resonates inside the instrument, and a woodwind instrument might be made of brass (see the saxophone for example).
A way to alter the timbre of a cylindrical duct would be to change the acoustic impedance at the bell. For that, the concept of Electroacoustic Resonator, developped in the lab, is likely to allow real-time modification of acoustic impedance at a certain position along the duct.
This project aims at developing a scaled prototype of an active wind instrument, made of an archetypal cylindrical hollow duct with straight terminations, around which a ring of active electroacoustic absorbers will be designed to control the acoustic impedance and alter the timbre of the duct. The work will consist in:
perform FEM simulations of the proposed design and identify potential control settingsdevelop an experimental prototype of wind-instrument “scale model”perform acoustic assessment of the electroacoustic resonators as well as of the whole “wind instrument”
Profile
Electrical EngineeringMechanical EngineeringMicroengineering, PhysicsRequirements
ElectroacoustiqueAudio engineering lecturesLearning outcomes
Acoustic instrumentationCOMSOL multiphysicsSpeedgoat programming (Matlab/Simulink)Content
COMSOL simulation (50%),design, including control (25%)acoustic assessment (25%)

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, Electrical Engineering, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Supervisor,

The spherical sound source is known to be an optimal geometry, due to the limited scattering effects over the sphere when compared to other geometries (circular or rectangular planar pistons for instance). However, this shape is impossible to realize with conventionnal loudspeakers, apart from some approximations (the dodecahedron sources, used for room acoustics measurements, among others).
The Corona Discharge (CD) principle has been demonstrated to allow achieving linear acoustic flow velocity source without relying on an intermediate membrane, presenting an almost perfect electroacoustic transduction with an extremely sharp impulse response. The CD loudpeaker generally consists of two electrodes, a “corona” electrode of extremely thin size (array of wires, needles, etc.) put at a sufficiently high voltage to ionize the surrounding medium particles, and a “collector” electrode, conductor of larger size than the “corona” connected to the ground, attracting the ions while being sufficiently transparent to particle streams (eg. a metallic grid).
With this configuration, the CD loudspeaker presents a combination of two intrinsic sources: a monopolar “Heat” source, due to the local heat exchanges occurring in the ionization process, and a dipolar “Force” source, resulting from the electrostatic force accelerating the charged particles (and the surrounding medium) back and forth around the transducer. This transducer has been proven to be an ideal flow velocity source, and a recent PhD thesis proposed a detailed model of the transducer that can serve now for further optimization.
Besides the absence of membrane, this arrangement allows imagining various geometries that are not possible with membrane-based loudspeakers. Especially, it could be the ideal transducer to achieve the so-called “pulsating sphere”, an ideal source that is generally used as a model for omnidirectional sources, but that cannot be realized with conventional types of loudspeakers.
The proposed semester project intends to work on identifying and eventually implementing a purely omindirectional CD loudspeaker configuration, and if time allows, extrapolate towards other types of canonical electroacoustic sources (cylidrincal, linear antenna, circular antenna, etc.).
The project will consist in the development of a COMSOL (and/or analytical) model of the transducer(s), that will serve for geometry optimization. After defining a set of optimal geometries, one or several prototype(s) will be constructed and tested in anechoic conditions.

ContentCOMSOL and/or Matlab simulationsElectroacoustic measurements
PrerequisiteBA5-Electroacoustiqueor MA1-Audio Engineering
Additional referenceLink to a tutorial video

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, Electrical Engineering, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Rahim Vesal – Role: Supervisor,

Nonlinear or time-varying systems have established a new paradigm in wave engineering as they push the limits of conventional materials. Nevertheless, the harmonic frequencies generated by the nonlinearity or the temporal modulation of the medium properties make it challenging to characterize their scattering properties. Matched sources minimize such constraints since the incident field is known (any reflection at the source is absorbed), thus reducing the number of unknowns to be resolved in the system.
Although extensively used in electromagnetic, matched sources are not widely developed in acoustics due to the highly reflective nature of electrodynamic speakers. To overcome this limitation, the matching condition can be achieved by active control. By sensing the pressure in front of a closed-box loudspeaker and applying a feedback control law that assigns a specific current, the effective impedance of the loudspeaker can be changed and designed, thus enabling impedance matching.
The project will consist of:
a quick literature review on acoustic active control and matched sourcesthe development of a numerical scheme (FDTD / Simulink) to realize an acoustic matched sourcethe implementation of an experimental set-up to characterize the performance of the electrodynamic matched source at multiple harmonicsthe extension of the study to a broadband-matched source, by switching the electrodynamic source type to an actively controlled corona discharge actuatorProfile: Electrical engineering, Micro-engineering, Physics, Mechanics
Prerequisites: Acoustics, Electroacoustics
Learning outcome: linear active control for acoustics, acoustic measurement
Context: Theory/Physical Simulations (40%), measurement (60%)

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, Electrical Engineering, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Rahim Vesal – Role: Supervisor,

The Electroacoustic Absorber (EA) concept relies on a closed-box loudspeaker, the membrane of which can be assigned a prescribed acoustic impedance thanks to a microphone-based active feedback control. Once set, the acoustic performance of the EA is usually assessed in an impedance tube under plane waves, following ISO 10534-2 standard with two microphones.
Usually, the control parameters of the EA are identified based on a physical model, assuming the knowledge of the loudspeaker’s electroacoustic parameters. However, the identification of these parameters are usually biased, and the environmental conditions (temperature, humidity, etc.) might also change, and this might affect the performance of the concept. But if the controller allows adjusting control parameters on-the-fly, the EA could be made self-adjustable and more robust to these changes. The project proposes to investigate possible means to turn the “fixed” EA into a more adjustable one, thanks to advanced signal processing techniques.
After being initiated to the EA concept and to Speedgoat/Simulink programming, the student will investigate the means to make the controller parameters self-adjustable to the environment, in a view to optimizing the achieved acoustic impedance with respect to a targetted one.
The project will consist of:
literature review on Electroacoustic Absorbers and on optimizationdevelopment of a model (Simulink) of an EAdevelopment of signal processing techniques to self-adapt the EA control to unknown and/or time-varying parametersexperimental set-up to characterize the performance of the EA in an impedance tube, with time-varying conditionsProfile: Electrical engineering
Prerequisites: Electroacoustics (optionally: optimization)
Learning outcome: active control in acoustics, acoustic measurement, Speedgoat programming
Context: Theory/Physical Simulations (30%), measurement (30%), Matlab programming (40%)

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, SSC, Electrical Engineering, Learning & Decision Systems, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Louis Léonard Marie Larcher – Role: Student, Qin Liu – Role: Supervisor,

The electroacoustic resonator concept developed by the LWE-Acoustic Group, employs a conventional (closed-box) loudspeaker, connected to one ore more microphone(s) through a real-time digital control unit allowing assigning target acoustic impedance through velocity control. The current solution relies on a massive laboratory-grade control hardware (Speedgoat real-time controller), that is not well suited for fast and modular demos. The aim of this project is to develop a stand-alone solution, relying on a light and mobile real-time controller, that meets the Electroacoustic Resonator specifications (input and output dynamic ranges, “real-time” processing meeting the targeted frequency range, including ADCs and DACs, etc.).

More specifically, the student will follow an ongoing project in the lab, aiming at developing a set of electroacoustic resonators to damp ventilation noise in a cooktop hood system. The work will consist in:Following the experimental assessment of the baseline ventilation noiseFollowing the development of the corresponding electroacoustic resonator solution on the Speedgoat hardwareAssisting in designing power amplifiers and microphones/microphone conditionersIdentifying the hardware solution for the final prototypePorting the identified processing on the new hardwareTests in laboratory conditions

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, SEL, Electrical Engineering,

Memberships:

Hervé Lissek – Role: Owner, Tim Tuuva – Role: Supervisor, Mathieu François,Mathieu Francois Padlewski – Role: Supervisor,

The acoustic metacrystal developed at LWE-Acoustic Group consists of multiple Electroacoustic Resonators (meta-atoms), arranged along an acoustic waveguide (a squared cross-section duct). The dynamics of the whole system can then be modified through multi-channel active control: a local control, aiming at adjusting the individual self-response to acoustic disturbance (for example to assign them the exactly same acoustic impedance), and a distributed control, aiming at modifying the interactions among successive atoms, as atoms do in a natural crystal. Thanks to this unique tool, it is possible to explore quantum phenomena with an audible device.
The prototype is expected to be an educative tool to teach quantum physics to EPFL students. The objective of the project is to replicated the current acoustic metacrystal prototype, with the same hardware (Speedgoat multichannel real-time controller, loudspeakers and microphones including all required electronics) and develop a user-friendly interface to allow changing the control parameters on the fly and, in the same time, perform acoustic measurements to explore the exotic properties of the device.
The project will the consist in:
learning the acoustic metacrystal functionalitiesreplicate the hardwaredesign and develop the intuitive user’s interface for educational purpose

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Electrical Engineering, Quantum Technology,

Memberships:

Hervé Lissek – Role: Owner, Mathieu François,Mathieu Francois Padlewski – Role: Supervisor,

The Corona Discharge (CD) principle has been demonstrated to allow achieving linear acoustic flow velocity source without relying on an intermediate membrane, presenting an almost perfect electroacoustic transduction with an extremely sharp impulse response. The CD loudpeaker generally consists of two electrodes, a “corona” electrode of extremely thin size (array of wires, needles, etc.) put at a sufficiently high voltage to ionize the surrounding medium particles, and a “collector” electrode, conductor of larger size than the “corona” connected to the ground, attracting the ions while being sufficiently transparent to particle streams (eg. a metallic grid).
With this configuration, the CD loudspeaker presents a combination of two intrinsic sources: a monopolar “Heat” source, due to the local heat exchanges occurring in the ionization process, and a dipolar “Force” source, resulting from the electrostatic force accelerating the charged particles (and the surrounding medium) back and forth around the transducer. This transducer has been proven to be an ideal flow velocity source, and a recent PhD thesis proposed a detailed model of the transducer that can serve now for further optimization.
Besides the absence of membrane, this arrangement allows imagining various geometries that are not possible with membrane-based loudspeakers. However, the absence of membrane imposes a “high-pass” behaviour for this sound source in the far field, making it almost inneffective as a “direct-radiation” sound source.
The proposed semester project intends to work on identifying and eventually implementing CD loudspeaker configurations improving the sound radiation efficacy, especially presenting a meaningful band-pass behaviour over a relatively large frequency band in the audible range, and eventually improving the radiated sound power level of the source.
The project will consist in the development of a COMSOL (and/or analytical) model of the transducer, that will serve for geometry optimization. After defining a set of optimal geometries, one or several prototype(s) will be constructed and tested in anechoic conditions.

ContentCOMSOL and/or Matlab simulationsElectroacoustic measurements
PrerequisiteBA5-Electroacoustiqueor MA1-Audio Engineering
Additional referenceLink to a tutorial video

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, Electrical Engineering, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Rahim Vesal – Role: Supervisor,

The Corona Discharge (CD) principle has been demonstrated to allow achieving linear acoustic flow velocity source without relying on an intermediate membrane, presenting an almost perfect electroacoustic transduction with an extremely sharp impulse response. The CD loudpeaker generally consists of two electrodes, a “corona” electrode of extremely thin size (array of wires, needles, etc.) put at a sufficiently high voltage to ionize the surrounding medium particles, and a “collector” electrode, conductor of larger size than the “corona” connected to the ground, attracting the ions while being sufficiently transparent to particle streams (eg. a metallic grid).
With this configuration, the CD loudspeaker presents a combination of two intrinsic sources: a monopolar “Heat” source, due to the local heat exchanges occurring in the ionization process, and a dipolar “Force” source, resulting from the electrostatic force accelerating the charged particles (and the surrounding medium) back and forth around the transducer. This transducer has been proven to be an ideal flow velocity source, and a recent PhD thesis proposed a detailed model of the transducer that can serve now for further optimization.
The absence of membrane allows for a fully transparent sound source, which could be an asset in certain circumstances: for example, most of the headphones/earphones owner suffer from the so-called “occlusion effect” that is likely to create a low-pass effect on the sound rendering. With a fully open configuration, the rendered sound is likely to be more natural. But the price to pay is a strong leakage of sound towards the environement, which needs to be taken in consideration and worked around.
The proposed semester project intends to work on simulating and eventually implementing a fully open CD headphone, and testing it on an artificial ear facility.
The project will consist in the development of a COMSOL (and/or analytical) model of the transducer(s), that will serve for geometry optimization. After defining a set of optimal geometries, one or several prototype(s) will be constructed and tested in anechoic conditions.
Content
COMSOL and/or Matlab simulationsElectroacoustic measurementsPrerequisite
BA5-Electroacoustiqueor MA1-Audio Engineering

Project Creator:

Hervé Lissek ([email protected])

Units Involved:

Laboratoire d''ingénierie des ondes (LWE)

Tags:

Master, Bachelor, SEL, SGM, SMT, Electrical Engineering, Mechanical Engineering, Microengineering,

Memberships:

Hervé Lissek – Role: Owner, Han Miao – Role: Supervisor, Emré Tokay – Role: Student,