Search

National audience; La reproduction binaurale a pour but de recréer une scène sonore réaliste aux oreilles de l’auditeur, à l’aide d’un casque audio. Malheureusement, plusieurs études montrent une baisse de l’externalisation lorsque des sujets écoutent des contenus binauraux avec HRTFs non-individualisées (les sources sonores sont perçues à l’intérieur de la tête, alors qu’elles devraient être localisées à l’extérieur de la tête).La présente étude examine si de larges mouvements de tête couplés à un système de ”head-tracking” peuvent compenser cette baisse d’externalisation. Pour chaque présentation, un stimulus (voix d’homme) était présenté au casque à divers azimuts, avec différentes HRTFs non-individualisées pour le rendu binaural. Le head-tracker pouvait être soit actif, soit inactif, et les sujets devaient soit bouger la tête, soit rester immobiles. Après chaque présentation, les sujets devaient évaluer le degré d’externalisation perçu. Les résultats ont montré que de larges mouvements de tête peuvent considérablement améliorer l’externalisation, et que cette externalisation persiste quand bien même le sujet a cessé de bouger la tête. Ces résultats ont été obtenus avec des sujets experts et naïfs.Une autre expérience a été menée, dans laquelle les rôles étaient inversés : les sujets gardaient la tête fixe tandis que les sources sonores tournaient automatiquement autour d’eux. Les résultats ont montré que de tels mouvements de sources peuvent aussi améliorer l’externalisation, mais dans une moindre mesure que les mouvements de tête.

International audience; In order to facilitate the development of wrist-worn vibrotactile devices, detailed knowledge about how vibrations are perceived by the users is needed. In particular, perceptual thresholds in amplitude are really important. Thresholds have been measured in the literature for other areas of the body, but given the variability reported between areas (shape of the threshold curve, position of maximum sensitivity), thresholds on the wrist can not be inferred from previous measurements and must be measured. The amplitude thresholds for vibrations normal to the skin surface were evaluated on 28 participants, with a three interval forced choice method. They were measured for 7 frequencies that are classical in the literature about vibrotactile perception (25, 40, 80, 160, 250, 320, and 640 Hz). The classical U-shape of the amplitude-threshold curve is observed, with a maximum sensitivity at around 160 Hz, which differs from other body areas, but confirms recent results obtained for vibrations parallel to the skin surface of the same body area. The sensitivity thresholds of vibrotactile signals appear to be in the micrometer range.

This chapter concerns concepts of adaption in a binaural audio context (i.e. headphone-based three-dimensional audio rendering and associated spatial hearing aspects), considering first the adaptation of the rendering system to the acoustic and perceptual properties of the user, and second the adaptation of the user to the rendering quality of the system. We start with an overview of the basic mechanisms of human sound source localisation, introducing expressions such as localisation cues and interaural differences, and the concept of the Head-Related Transfer Function (HRTF), which is the basis of most 3D spatialisation systems in VR. The chapter then moves to more complex concepts and processes, such as HRTF selection (system-to-user adaptation) and HRTF accommodation (user-to-system adaptation). State-of-the-art HRTF modelling and selection methods are presented, looking at various approaches and at how these have been evaluated. Similarly, the process of HRTF accommodation is detailed, with a case study employed as an example. Finally, the potential of these two approaches are discussed, considering their combined use in a practical context, as well as introducing a few open challenges for future research.

International audience

International audience; This paper addresses the issue of controlling monodic pitch in digital musical instruments (DMIs), with a focus on in- struments for which the pitch needs to be played with accu- racy. Indeed, in many cultures, music is based on discrete sets of ordered notes called scales, so the need to control pitch has a predominant role in acoustical instruments as well as in most of the DMIs. But the freedom of param- eter mapping allowed by computers, as well as the wide range of interfaces, opens a large variety of strategies to control pitch in the DMIs. Without pretending to be ex- haustive, our paper aims to draw up a general overview of this subject. It includes: 1) a review of interfaces to pro- duce discrete and/or continuous pitch 2) a review of DMI maker strategies to help the performer for controlling eas- ily and accurately the pitch 3) some developments from the authors concerning interfaces and mapping strategies for continuous pitch control 4) some comparisons with acous- tical instruments. At last, a Max/MSP patch -publically available- is provided to support the discussion by allow- ing the reader to test some of the pitch control strategies reviewed in this paper.

International audience

International audience; String excitation by the tangent in the clavichord is a unique mechanism. The tangent, keeping in contact with the string after the initial strike, continuously controls the string tension. Four main flexible subsystems are considered in the clavichord: the tangent/key subsystem, the string subsystem, the bridge-soundboard subsystem, and the string damper subsystem. A modal description of the dynamics of these subsystems is proposed. Parameters of the subsystems are estimated on a copy of a historical instrument by Hubert (1784). The different subsystems and their couplings are modeled using a modal Udwadia–Kalaba formulation. The string-tangent interaction is modeled via the intermittent contact dynamics, using the Kirchoff–Carrier string model. Realistic string, soundboard, and tangent motions are obtained using a time-domain synthesis scheme that computes the dynamics of the uncoupled subsystems and the constraints resulting from coupling between them. Simulated motions of the model in response to a driving force on the key are analyzed. The results are consistent with experimental measurements and published data on the dynamics of the clavichord. The model is able to reproduce the main acoustic features of the instrument: force on the key for intonation control, key velocity for dynamic nuances control, and constant spectral slope for varying dynamic nuances.

International audience; Auralization technology has reached a satisfactory level of ecological validity, enabling its use in architectural acoustic design. Only recently have the actual uses of auralization in the consulting community been explored, resulting in the identification of a variety of uses, including (1) to present to clients, (2) to test design ideas, (3) as a verification tool, (4) as a verification tool, (5) as a marketing tool, and (6) to improve internal company discussions. Taking advantage of methodologies from ergonomics research, the present study investigates effective uses through the observation of a collaboration project between an acoustic research team and an acoustic consultant, as a case study. Two spaces have been auralized in the context of the conception of a new skyscraper during the design phase of the project. The two spaces faced different problematics: an Atrium for which three different acoustic treatment options were suggested and experienced through multi-modal auralizations and audio-only auralizations of an Auditorium where an intrusive noise was to be acoustically treated. The ergonomic observation and analysis of this project revealed key impediments to the integration of auralization in common acoustic design practices.

Cet article décrit le projet de recherche OrJo 2009-2012 (Orchestre de Joysticks) qui associe quatre structures, PUCE MUSE, le LAM (UPMC), le LIMSI (CNRS, associé à l’UPMC et à l'Université Paris-Sud), et 3Dlized, autour de quatre grands objectifs :\ud 1. réaliser quatre versions du logiciel plateforme pour s'adapter aux différents usages,\ud 2. proposer une collection d’instruments virtuels sonores et visuels,\ud 3. améliorer la représentation graphique des instruments virtuels,\ud 4. pratiquer, échanger, et conserver un répertoire sur la Méta-Librairie.\ud Ce projet interroge plusieurs usages nouveaux comme la pratique en orchestre d'instruments virtuels, l'échange et la transmission de partitions interactives pour ces orchestres, l'apport du relief pour la musique visuelle. Il évoque aussi les développements d'instruments virtuels comme les instruments de synthèse vocale du LIMSI et les instruments par modèles physiques, modèles topologiques et modèles statistiques de l'équipe Lutheries acoustique musique (LAM).\ud Cet article est aussi une invitation à utiliser la plateforme Meta-Mallette [12] via son SDK (libre) et le site d'echanges Meta-Librairie.

International audience; Le découplage entre le geste de l'instrumentiste et le son produit par son instrument dans les musiques électroniques a entrainé une reconsidération totale de la notion d'instrument [5] et de nouvelles pratiques pour la facture d'instruments numériques [1]. Pour concevoir des instruments numériques riches, le facteur est souvent amené à réaliser des connexions complexes et des réglages très fins, afin de retrouver à la fois un son vivant et une synesthésie entre le geste et le résultat acoustique qui permet in fine de pouvoir anticiper ou "chanter" ce que l'on joue. Il est possible d'obtenir un couplage riche et intuitif en procédant empiriquement au câblage et aux courbes de transfert entre toutes les variables d'entrée et de sortie, mais cette opération fastidieuse se fait souvent au détriment de la modularité informatique, qui permettrait de changer d'interface ou de synthèse en un simple clic. Les auteurs, qui ont déjà présenté le concept de "Modèles Intermédiaires Dynamiques" (MID) pour répondre à ce problème, présentent dans cet article leurs derniers développements concernant le contrôle de macro-formes par des processus génératifs. L'utilisation d'un automate cellulaire connu sous le nom de "Jeu de la Vie" est plus particulièrement présentée ici.

International audience; Mapping is one of the most important aspects of software instruments design. We call "mapping" the relation defined between the parameters from hardware interaction devices, and those of the process to be controlled. For software instruments, this relation between the user's gestures and synthesis engine parameters has a decisive role in resulting ergonomics, playability and expressive possibilities of the system. The authors propose an approach based on a modular software design inspired by a multidisciplinary study of musical instruments and their playing.

International audience; Lors du développement d'instruments logiciels, la manière dont les données issues des interfaces gestuelles sont mises en relation avec les paramètres nécessaires au contrôle de la synthèse, ou mapping, a un rôle décisif dans l'ergonomie, la jouabilité et les possibilités expressives du dispositif. Les auteurs proposent une approche s'appuyant sur une conception logicielle modulaire.

International audience; Studying the directivity of musical instruments such as pipe organs is challenging because of their great size and because they typically cannot be moved from the locations where they are built to the laboratory. This study compares the radiation pattern of organ buffets under different excitation conditions, using a 19th century French pipe organ. We investigate the positive and the great organ independently due to their spatial separation and size. The excitation strategies were comprised of cylindrical and an omnidirectional electroacoustic sources. Measurements were carried out on each organ buffet along a horizontal line spanning the width of the church. Results in octave bands are shown and compared with a discussion on possible causes for observed differences. Significant variations in directivity were observed for the 2 kHz to 4 kHz octave-band regions where scattering from pipes is expected to have a predominant effect, while little variation from omnidirectional was observed for lower frequency bands for both sources.

International audience

International audience

International audience; Significant architectural modifications were made to Notre-Dame de Paris between the 1230s and the middle of the 14th century despite the completion of the central volume in the 1220s. These changes include the restructuring of the windows and attics along the clerestory and the gradual outward expansion of the building. Mainly, the transept arms were widened and side chapels were constructed between the buttresses along the perimeter of the cathedral. This study explores the acoustic effect the modifications to the transept and side chapels may have had on sound produced in the cathedral.

International audience; While the radiation pattern of small instruments can be measured in anechoic chambers, it is challenging to study the directivity of large, unmovable instruments such as pipe organs associated with cultural historic sites. The soundfield outside an organ buffet is the result of the convolution of the sound of the pipe with the surrounding scattering system, the enclosing cavity of the instrument, and the apertures between the pipes in the façade of the instrument. The directivity of a small, proxy organ buffet consisting of a plywood enclosure with PVC pipe scatterers was studied in an anechoicchamber using omnidirectional and cylindrical loudspeaker sources. This proxy buffet grants access to modeling parameters such as the pipe spacing and density. These laboratory measurements were then compared to in situ, free-field directivity measurements of the same sources inside a real positive organ buffet of comparable size and shape located inside a church.

International audience; The primary construction of the Cathédrale Notre-Dame de Paris began in the spring of 1163 CE and continued into the mid-14th century. Two decades after construction efforts commenced, the pre-existing liturgical community in Paris began its occupancy inside the incomplete building. To better understand the interrelationship between the occupants and the cathedral, the acoustical characteristics of key moments in the cathedral’s early development in the 12th and 13th centuries are examined using geometric acoustic modeling. Room acoustic parameters of the sacred regions of the cathedral are examined with attention to the experiences of active participants within the chancel of the cathedral.

International audience; In the middle of the 14th century, the recently constructed Great Clementine Chapel of the Palais des Papes had a flourishing reputation for the composition and interpretation of polyphonic singing in the emerging Ars Nova musical style. In modern times, the space is still employed for musical performances. However, the acoustic conditions between the two periods vary greatly, and as such, can be expected to have an impact on vocal performances. As part of the IMAPI and PHE projects, the impact of the acoustics of the Great Clementine Chapel on the performance of a conducted vocal ensemble specializing in medieval music was examined for these two periods. A numerical simulation of the medieval acoustics was developed, based on a calibrated geometrical acoustics model of the modern-day chapel which was then regressed in time to a historically informed medieval state. Experiments were carried out with singers performing repetitions of several pieces in a Virtual Acoustic Environment (VAE) using close-mics and headphone renderings. Recorded performances were analyzed using various metrics, with objective results paired with questionnaires acquired for each VAE condition. Preliminary analysis of these results is presented in this study.

International audience; When we think about great architectural achievements in European history, such as ancient amphitheatres or Gothic cathedrals, their importance is strongly tied to their acoustic environment, an intangible consequence of the space's tangible construction and furnishings. We present an overview of the "Past Has Ears" project and current progress, exploring how the acoustics of heritage spaces can be documented, reconstructed, and experienced for spaces existing, altered, or destroyed.

International audience; Performative speech synthesis is a new avenue for expressive speech and singing analysis andsynthesis. At the confluence between new interfaces for musical expression, voice signal processingand acoustic phonetics, two vocal instruments are presented. Cantor Digitalis, a formant synthesizerusing a spectral glottal model, is controlled using bimanual gestures on a graphic tablet. Voks is basedon sampling and vocoding. It is controlled by bimanual gestures (graphic tablet or theremin) andbiphasic syllabic sequencing. Prosodic units and their organization for vocal instruments are discussedalong three lines: chironomic stylization of intonation and vocal effort, biphasic syllabic rhythmcontrol. Musical, pedagogical and clinical applications of this new organology of voice are discussed.; La synthèse vocale performative offre une nouvelle approche pour l’analyse et la synthèse de la paroleexpressive et du chant. Deux instruments vocaux, à la confluence entre nouvelles interfaces pourl’expression musicale, traitement du signal vocal et phonétique acoustique, sont présentés. CantorDigitalis est un synthétiseur à formants muni d’un modèle spectral de la source glottique, piloté pardes gestes bimanuels sur une surface tactile et un stylet. Voks est un vocodeur piloté par des gestesbimanuels, sur une surface tactile et un stylet, ou par un thérémine, augmenté d’un contrôle biphasiquedu séquencement syllabique. Les unités prosodiques, leur organisation et leur contrôle pour laconstruction des instruments vocaux sont discutés suivant trois lignes : la stylisation chironomique del’intonation et de l’effort vocal, le contrôle biphasique du rythme syllabique. La conclusion évoqueles applications musicales, pédagogiques, voire cliniques de cette nouvelle organologie de la voix.

Simulating the acoustics of destroyed or altered amphitheaters, cathedrals, and other architectural sites re-creates their sonic grandeur.

With his work on the Fogg Art Museum and Boston Symphony Hall between 1895 and 1900, Wallace C. Sabine laid a foundation for the field of architectural acoustics as a science. Prior to that, architects employed various quantifiable notions in acoustic design. Previous studies have reviewed metric guidelines based on the directivity of the human voice, which was utilized in at least 11 rooms in pre-Sabine times. Others studies have reviewed pre-Sabine design guidelines that were based on the quantification of the perception threshold between direct sound and first order reflections and which were followed in several rooms with acoustical performance needs. As the first studies concerned the direct sound and the second set concerned first order reflections, this study reviews opinions and knowledge regarding the later part of the acoustic response, also known as reverberation, during the 19th century. This effort brings to light a room acoustic design evolution showing why concert halls at the end of the 19th century mainly had surface finishes of wood and plaster as well as limited ceiling heights. While not equal to Sabine's thoroughness and completeness, numerous early experimental and theoretical reverberation approaches were found to hint at similar notions with both qualitative and quantitative efforts.

Every individual perceives spatial audio differently, due in large part to the unique and complex shape of ears and head. Therefore, high-quality, headphone-based spatial audio should be uniquely tailored to each listener in an effective and efficient manner. Artificial intelligence (AI) is a powerful tool that can be used to drive forward research in spatial audio personalization. The SONICOM project aims to employ a data-driven approach that links physiological characteristics of the ear to the individual acoustic filters, which allows us to localize sound sources and perceive them as being located around us. A small amount of data acquired from users could allow personalized audio experiences, and AI could facilitate this by offering a new perspective on the matter. A Bayesian approach to computational neuroscience and binaural sound reproduction will be linked to create a metric for AI-based algorithms that will predict realistic spatial audio quality. Being able to consistently and repeatedly evaluate and quantify the improvements brought by technological advancements, as well as the impact these have on complex interactions in virtual environments, will be key for the development of new techniques and for unlocking new approaches to understanding the mechanisms of human spatial hearing and communication.

International audience; Through a review of the current literature, this chapter defines a methodology for the analysis of HRTF localisation performance, as applied to assess the quality of an HRTF selection or learning program. A case study is subsequently proposed, applying this methodology to a cross-comparison on the results of five contemporary experiments on HRTF learning. The objective is to propose a set of steps and metrics to allow for a systematic assessment of participant performance (baseline, learning rates, foreseeable performance plateau limits, etc.) to ease future inter-study comparisons.

International audience

International audience; Temps de discussion à l'issue de la session Enseignement de l’Acoustique - méthodes, outils et pratiques.

International audience; Afin de transmettre des informations à une personne de manière vibrotactile, l’usage de transducteurs électrodynamiques est souvent privilégié car il présente une grande dynamique et une large bande passante. Pour caractériser ces systèmes, la réponse en fréquence est classiquement donnée pour un dispositif non couplé (donc sans pression sur la partie mobile) au système excité (peau de l’utilisateur). Or, lors de cette association entre la peau et le transducteur, la contrainte mécanique exercée modifie le comportement du transducteur. Il est donc nécessaire de connaître la réponse en fréquence lorsque le dispositif est couplé. Différents tests ont montré que les méthodes utilisant un capteur de force ne nous permettent pas d’estimer correctement la pression du transducteur sur la peau (surface de contact trop faible, position fixe du bras difficile à tenir...). Les technologies alternatives (transducteur hydraulique, pneumatique, ressorts comprimés…) deviennent vite invasives ou moins résilientes. De plus, ce couplage, en déformant la peau et ses mécanorécepteurs, altère la perception des signaux vibrotactiles. L’intérêt de caractériser le transducteur chargé est donc double, il permet de mesurer la charge appliquée au transducteur (qui modifie le signal généré) mais aussi la charge appliquée sur la peau (qui modifie le signal perçu). Nous proposons une estimation de la pression exercée sur le transducteur à partir de la mesure de son impédance électrique. Pour cela, un banc d’essais a été spécifiquement développé et a pu être validé. Il permet d’établir des liens entre l’évolution de la courbe d’impédance du transducteur (décalage de la fréquence de résonance, évolution de l’amplitude de l’impédance...) et la charge mécanique appliquée. Une modélisation électromécanique du transducteur permet alors d’interpréter ces observations. L'exploitation de cette modélisation améliore la répétabilité des couplages vibreur-peau. Entre autres, la pression exercée sur la peau peut être ajustée en mesurant la fréquence de résonance du transducteur.

International audience; Lors de fouilles dans les années 2002-2005 autour de la pyramide du pharaon Nebkheperrê Antef, située sur la rive ouest de Louxor, dans la nécropole de Dra’ Abou el-Naga, les archéologues ont trouvé trois harpes. L’une d’entre elles est dans un état de conservation exceptionnel en dépit de son aspect fragmentaire au moment de son exhumation. Après restauration, le corps de l’instrument taillé dans une seule pièce de bois est entier et même une partie de sa table d’harmonie en peau a survécu aux outrages du temps. La rareté de ce type de vestige a incité à mettre en place une étude interdisciplinaire pour essayer de tirer le maximum d’informations de l’objet, dont une analyse sur ses propriétés acoustiques. La qualité sonore de cette harpe de l’Égypte ancienne ne pouvant pas être appréhendée à partir du vestige archéologique, une copie de l’instrument a été réalisée par une luthière qui avait déjà une expérience dans la reconstruction de cordophones antiques. Elle est partie d’un relevé archéologique précis des dimensions de l’original et a tenu compte des différents types de bois identifiés pour le corps, le cordier et les « chevilles ». L’analyse vibratoire et acoustique de cette copie a permis, d’une part, de mieux saisir l’expertise des artisans de l’Antiquité qui ont conçu ce cordophone vers 1600 avant notre ère et, d’autre part, d’ouvrir une réflexion sur l’accordage de l’instrument. De plus, une comparaison avec l’acoustique de harpes actuellement fabriquées et jouées en Afrique Centrale montre des propriétés proches, mettant en lumière une conception qui présente des caractéristiques similaires, malgré des choix différents de facture.

International audience; Les similarités entre les signaux audios et vibrotactiles nous amènent naturellement à comparer nos capacités à les appréhender (discrimination fréquentielle, courbes d’isosensiblité…). L’utilisation de technologies (transducteurs électrodynamiques) proches dans les deux domaines renforce les occasions d’effectuer des comparaisons. Lors de la numérisation des signaux audio, des quantifications sur 16, 24 voire 32 bits sont classiquement réalisées. Ces niveaux de quantification permettent de limiter sa perceptibilité. Cependant, la quantité de données générée est proportionnelle au niveau de quantification. La numérisation sur un grand nombre de bits demande donc l’usage de matériels adaptés. La pertinence d’utiliser des hauts niveaux de quantification se pose donc régulièrement. Pour un même signal, toutes les variations (spectrale, d’amplitude…) de signal perceptibles au toucher le seront auditivement. En revanche, des variations perceptibles auditivement ne le sont pas forcément au toucher. On peut donc légitimement penser que des niveaux de quantifications plus faibles (8 voire 4 bits) pour un même signal pourraient être perceptibles en diffusion audio sans l’être sur le canal vibrotactile. Nous proposons d’évaluer les seuils de discrimination de signaux vibrotactiles quantifiés sur différents niveaux. Pour cela, nous utilisons un pot vibrant (dont le comportement est contrôlé par un vibromètre), et un protocole de tests perceptifs habituel issu de l’audio. L’expérience est ensuite reproduite avec des transducteurs électrodynamiques du marché (utilisés dans les applications vibrotactiles) en tant qu’excitateurs afin d’étendre les résultats. Les résultats montrent que le niveau de quantification des signaux vibrotactiles (comparés à ceux des signaux audio) peut être considérablement diminué sans que la différence soit perceptible au toucher. Ils permettent donc de penser que certaines technologies destinées aux systèmes embarqués (convertisseurs analogique-numérique de quelques bits par exemple) pourraient être utilisées en diffusion vibrotactile sans que cela perturbe la perception du signal, levant ainsi la contrainte du matériel adapté précédemment évoquée.

International audience; Nous étudions la sympathie en tant qu’élément organologique, phénoménologique et esthétique des cordophones à clavier (famille des clavicordes, clavecins, pianos), tant pour le timbre que pour les modes de jeu, les styles d’interprétation et l’écriture. Du point de vue organologique, différents systèmes d’étouffoirs existent pour certains types d’instruments, mais pas pour tous. Ils définissent, de façon permanente ou de façon intermittente sous le contrôle de l’instrumentiste, les cordes ou ensembles de cordes mises en vibration sympathique. Du point de vue phénoménologique, la sympathie présente tout une gamme d’effets sonores, qui s’apparentent à la réverbération et à la résonance. Du point de vue esthétique, la sympathie est soit statique, une donnée du timbre de l’instrument (par exemple pour le clavicorde ou les cordes aliquotes du piano), soit dynamique, contrôlée et évoluant dans le temps. Le contrôle dynamique de la sympathie dans la pédalisation est devenue un des éléments esthétiques du jeu pianistique au cours de l’évolution musicale. L’instrument comme corps résonant a pris de l’importance dans la musique contemporaine : le jeu dans l’instrument, la percussion, le chant dans l’instrument, le piano préparé sont autant de façons d’exploiter la résonance pour créer des sonorités, des modes de jeu, des figures musicales nouvelles. Même si les cordophones à claviers ne sont généralement pas considérés comme des instruments « à cordes sympathiques », c’est dans cette famille d’instruments que la sympathie a dépassé le stade d’élément statique du timbre instrumental pour devenir un élément de discours musical et de son interprétation, élément avec lequel on compose.

International audience; In honour of the International Year of Sound and for the 1 year memorial of the Notre-Dame cathedral fire, a team of researchers and sound engineers created a virtual reconstruction of a concert in the cathedral, using close-mic recordings made on 24-April-2013 of a performance of La Vierge by Jules Massenet. This reconstruction was carried out during the period of strict COVID-19 confinement. With 83 musicians, 6 singers and a 160person choir spatially distributed throughout the cathedral, the original performance offered a spatial composition highlighting the complex acoustics and interactions between source and listener positions. Individual tracks were convolved with spatial room impulse responses, created from a calibrated geometrical acoustic simulation. Several listening positions were binaurally rendered, along with an artistic mix created by a sound engineer, offering a unique 3d-audio experience approaching the reality of the moment in the past. Distributed on-line, the website presenting this virtual reconstruction has been visited from users around the world. A listener opinion survey was included on the website. This paper presents an overview of the source material, the production workflow, and the challenges of realising such a production during confinement. Finally, an overview of visitor statistics and survey results is presented, providing insight into the reception of the virtual recreation and the interest towards future productions.

International audience

International audience; This research aims to build a prosodic boundary prediction model for improving the naturalness of Vietnamese speech synthesis. This model can be used directly to predict prosodic boundaries in the synthesis phase of the statistical parametric or end-to-end speech systems. Beside conventional features related to Part-Of-Speech (POS), this paper proposes two efficient features to predict prosodic boundaries: syntactic blocks and syntactic links, based on a thorough analysis of a Vietnamese dataset. Syntactic blocks are syntactic phrases whose sizes are bounded in their constituent syntactic tree. A syntactic link of two adjacent words is calculated based on the distance between them in the syntax tree. The experimental results show that the two proposed predictors improve the quality of the boundary prediction model using a decision tree classification algorithm, about 36.4% (F1 score) higher than the model with only POS features. The final boundary prediction model with POS, syntactic block, and syntactic link features using the LightGBM algorithm gives the best F1-score results at 87.0% in test data. The proposed model helps the TTS systems, developed by either HMM-based, DNN-based, or End-to-end speech synthesis techniques, improve about 0.3 MOS points (i.e. 6 to 10%) compared to the ones without the proposed model.

International audience; This paper introduces an interface that enables the real-time gestural control of intonation in phrases produced by a vocal synthesizer. The melody and timing of a target phrase can be modified by tracing melodic contours on the touch-screen of a mobile tablet. Envisioning this interface as a means for nonnative speakers to practice the intonation of a foreign language, we present a pilot study where native and non-native speakers imitated the pronunciation of French phrases using their voice and the interface, with a visual guide and without. Comparison of resulting F0 curves against the reference contour and a preliminary perceptual assessment of synthesized utterances suggest that for both non-native and native speakers, imitation with the help of a visual guide is comparable in accuracy to vocal imitation, and that timing control was a source of difficulty.

International audience; Speech glottal flow has been predominantly described in the time-domain in past decades, the Liljencrants-Fant (LF) model being the most widely used in speech analysis and synthesis, despite its computational complexity. The causal/ anti-causal linear model (LF CALM) was later introduced as a digital filter implementation of LF, a mixed-phase spectral model including both anti-causal and causal filters to model the vocal-fold open and closed phases, respectively. To further simplify computation, a causal linear model (LF LM) describes the glottal flow with a fully causal set of filters. After expressing these three models under a single analytic formulation, we assessed here their perceptual consistency, when driven by a single parameter R d related to voice quality. All possible paired combinations of signals generated using six R d levels for each model were presented to subjects who were asked whether the two signals in each pair differed. Model pairs LF LM-LF CALM were judged similar when sharing the same R d value, and LF was considered the same as LF LM and LF CALM given a consistent shift in R d. Overall, the similarity between these models encourages the use of the simpler and more computationally efficient models LF CALM and LF LM in speech synthesis applications.

International audience; This article describes Mono-Replay, a software environment designed for sound animation. "Sound animation" in this context means musical performance based on various modes of replay and transformation of all kinds of recorded music samples. Sound animation using Mono-Replay is a two-step process, including an off-line analysis phase and on-line performance or synthesis phase. The analysis phase proceeds with time segmentation, and the set up of anchor points corresponding to temporal musical discourse parameters (notes, pulses, events). This allows, at the performance phase, for control of timing, playback position, playback speed, and a variety of spectral effects, with the help of gesture interfaces. Animation principles and software features of Mono-Replay are described. Two examples of sound animation based on beat tracking and transient detection algorithms are presented (a multitrack record of Superstition by Steve Wonder and Jeff Beck and Accidents/Harmoniques, an electroacoustic piece by Bernard Parmegiani). With the help of these two contrasted examples, the fundamental principles of "sound animation" are reviewed: parameters of musical discourse, audio file segmentation, gestural control and interaction for animation at the performance stage.

This study involved playing and listening (using recorded sounds) experiments to investigate how changes in soundpost length (for a fixed soundpost position) affect the perceptual qualities of the violin and what the threshold of change is below which players and luthiers do not perceive differences. A length-adjustable carbon fiber soundpost was employed. During the playing experiment, subjects played a provided violin on which the soundpost length was modified by the experimenter to find their optimal soundpost lengths. Then the experimenter varied the soundpost length randomly in ten trials within ±0.11 mm around their optimal lengths and asked subjects to always compare with the previous setting. The results showed that subjects' optimal soundpost lengths varied from 0.132 to 0.616 mm relative to the original length (53 mm), but subjects could not recognize length variation of 0.11 mm or less at above chance levels. During the listening experiment, subjects listened to 16 pairs of recordings through a computer interface and were asked, for each pair, whether the violin setup was the same or different. The results showed that subjects could differentiate soundpost lengths with a difference of about 0.198 mm at better than chance level.

International audience; The RUMBLE project aimed to quantify the effects of low sonic boom exposure on human responses, in situations representative of the daily life of European citizens, so as to inform policy, establish standards and develop protection concepts for a new generation of supersonic commercial aircraft that should emit a reduced but perceivable boom (a ‘low boom’) while flying overland. Two low boom simulators were affixed to the bedrooms’ windows of a house located on our university campus. The simulators allowed us to study indoor the participants’ responses to realistic “outdoor” booms. Testing took place in both the living room and kitchen because the booms caused different intensities of rattle noise in those two rooms. Participants performed a communication task, a working memory task, a drawing task, and an automatic evaluation task. They also had three mandatory rests. We used questionnaires to assess the booms’ perception and annoyance as well as their influence on participants’ mood. The booms were rated unpleasant, especially with rattle, and more disturbing during the rests than during the tasks. The booms made the rests less relaxing and impaired short-term cognitive performance. [This research was conducted as a part of Rumble - EU Grant Agreement 769896.]

International audience; This study examines the efficiency of a training protocol using a virtual reality application designed to accelerate individual’s selection of, and accommodation to, non-individualized HRTF profiles. This training introduces three elements to hasten audio localization performance improvement: an interactive HRTF selection method, a parametric training program based on active learning, and a relatively dry room acoustic simulation designed to increase the quantity of spatial cues presented. Participants rapidly selected an HRTF (≈5 min) followed by training over three sessions of 12 min distributed over 5 days. To study the impact of the room acoustic component on localization performance evolution, participants were divided into two groups: one acting as control reference, training with only anechoic renderings, the other training in reverberant conditions. The efficiency of the training program was assessed across groups and the entire protocol was assessed through direct comparisons with results reported in previous studies. Results indicate that the proposed training program led to improved learning rates compared to that of previous studies, and that the included room response accelerated the learning process.

International audience

International audience

Prior to Sabine’s work on the Fogg Art Museum and Boston Symphony Hall, several numerical guidelines had been developed and applied to the design of rooms with specific acoustic demands such as theatres, concert halls, and opera houses. Previous papers have discussed guidelines based on the following principles: voice directivity, which was employed in the design of at least 11 rooms, “echo theory”, which quantifies the perception threshold between direct sound and first order reflections in order to prevent echoes from occurring, aiding in the design of at least 7 rooms and leading to the first known use of an acoustic scale model, and notions of reverberation, which influenced the design of at least 14 rooms. This paper discusses three additional pre-Sabine numerical guidelines that were used in room acoustic design: (1) audience rake, (2) stage acoustics and proscenium design, and (3) length, width, and height ratios. The origin of these theories, as well as examples of rooms in which they were applied, are discussed and compared to current practices in room acoustic design.

International audience; Auralizations can be computed in a variety of ways as well as be rendered over different sound reproduction systems. They are used as a design tool in architectural projects and for fundamental studies on spatial perception and cognition, hence requiring reliability and confidence in the obtained results. This study assessed this reliability through auditory perception stability by comparing the perceived differences between two rendering systems for a given set of second-order Ambisonic auralizations: virtual loudspeaker binaural rendering over head-tracked headphones versus 32-loudspeaker rendering. Anechoic extracts of jazz pieces have been recorded and presented in various acoustic conditions over these two systems, evaluated on the following criteria: Readability, distance, listener envelopment (LEV), apparent source width (ASW), reverberance, and loudness. Results show that consistent significant differences between scene conditions are comparably perceived across the two systems. However, significant effects of the sound reproduction system were observed for ASW, LEV, and reverberance in some configurations. V

International audience; The goal of this research is to identify relations between the perceived qualities (timbre, playability, …)of a given individual musical instrument (as evaluated by players and listeners) compared to otherinstruments of the same family and the description of its structural and acoustic properties. Here, we arenot comparing the timbre of different instruments as explored through the evaluation and measurementof a single note of for example a clarinet versus a single note of a trumpet, but the timbre of this violin(over the whole register) compared to the timbre of that violin. Our aim is to identify the acousticphenomena that can account for these perceptual timbre differences, which could provide insight toviolin makers about how to empirically modify some construction parameters in order to reach somedesired sound qualities.In the light of these various perceptual tests conducted on violins and violas associated to verbal analysesof questionnaires, we will show that there is no strict mapping between the perceived timbre and audiodescriptors as well as vibro-acoustical measurements, due to the complexity of the concept of timbrethat induce a large variability between musicians: the lack of consensus between the musicians’evaluation criteria actually results from the diversity of interpretations when evaluating the timbre of aninstrument globally (across the whole register, for different dynamics and playing techniques, …) andassessing the subtle acoustic differences between instruments of the same family.

International audience; La voix humaine peut produire et imiter toutes sortes de sons. S’appuyant sur les fonctions physiologiques (respiratoires, phonatoires, articulatoires) qui lui permettent de parler et de communiquer, le chanteur apprend à ajuster et à contrôler finement les diverses parties de son organe vocal pour modeler le timbre dans l’expression musicale. Nous disposons aujourd’hui de puissants moyens d’analyse pour étudier le fonctionnement de l’appareil vocal et les différents niveaux de son contrôle. Il est important néanmoins de prendre aussi en compte le fait que le timbre vocal s'inscrit fortement dans une culture au sein de laquelle priment les effets perceptifs et les interactions plus ou moins conscientes qui existent entre les chanteurs et les auditeurs en vue d'un résultat global à atteindre.C’est dans ce contexte que nous analysons le lien entre geste vocal et timbre vocal en appuyant notre argumentaire sur diverses réalisations remarquables du timbre vocal mises en œuvre dans quelques chants traditionnels : les styles teshka et leka du chant féminin bulgare ; le yodel autrichien et le tahrir iranien ; trois styles de chants diphoniques : tuva, mongol, xhosa ; un chant polyphonique de Sardaigne et l’art vocalique des Aka de Centrafrique et des Bunun de Taïwan.Nous abordons la complexité de la notion de timbre vocal et de sa perception en différenciant d'une part les techniques vocales du chant soliste qui s'appuient sur la pratique de gestes phonatoires et articulatoires spécifiques, et celles que l'on peut remarquer dans les polyphonies qui consistent soit en contraste de timbres entre les voix individuelles, soit en fusion de timbres différenciés en vue d'un effet global spécifique interagissant avec la sensation de hauteur.Ces réflexions nous amènent aussi à questionner le croisement entre hauteur musicale, contenu vocalique et timbre.Retrouvez sur cette page du site de l'association Colophon - https://www.colophon.be/05-docu-la-voix - une version multimédia adaptée

International audience; The Bilbao project aimed at relating intrinsic characteristics of the materials (wood density and stiffness) and some geometric characteristics of the violin's constituent part (thicknesses of the plates) with the tonal qualities of the complete violins. To this end, six instruments were carefully built at the Bilbao making school: three instruments with normal backs, each paired with a pliant (thin), normal or resistant (thick) top; similarly, three with normal tops, each paired with a pliant, normal or resistant back. The two examples of normal top paired with normal back serve as a control. Wood for tops and backs were closely matched in density and sound speeds-all tops and backs from the same trees. Greater control was achieved by having all plates and scrolls cut by CNC routers, using the Huberman Stradivari model. The outside surface was not changed as the graduation was performed entirely on the inside surface. In addition, another six instruments were built by six established makers, following a similar procedure but with less constraints on the choice of wood (not the same trees as for the Bilbao set though a similar density was imposed) and on the graduation of the routed plates which was left totally free (but ended in the same range of thicknesses). Finally, another violin built with a very different profile but made by one of the established maker was added to the pool as an outlier. These thirteen violins were then evaluated by twenty players during a free categorisation task and by about 70 listeners (31 violin makers, 26 bow players and 15 others) during a listening test in the Bilbao conservatory auditorium. The tests show very large differences in terms of timbre, playability and volume between the violins, and these differences will be discussed in the light of their construction parameters.

International audience; Musical acoustics requires many notions in mechanics, acoustics, physics, signal processing, etc. Unfortunately, in the context of teaching in a Master of Science, the number of hours devoted to this kind of teaching is extremely limited and requires the development of appropriate teaching methods. In this presentation, we will present the approach we have implemented as part of the ATIAM Master's degree (Acoustics, Signal Processing and Computer Science applied to Music) by defining pedagogical objectives, developing pedagogical modalities and establishing evaluation sequences. In particular, in this teaching, we mix both Problem-Based Learning and Project-Based Learning with more traditional courses such as lectures and practical works. The idea is to encourage students to be autonomous in their learning (and therefore to encourage acquisitions for which the student has gaps) and to motivate them to carry out these acquisitions by proposing appropriate problem situations. Thus, we will present the global structure of the module built around the modelisation of musical instruments. It includes problem situations implemented to achieve the learning objectives around percussion, string and wind instruments, and a project dealing with radiation, sound recording and room acoustics. The students' results, since the implementation of this course, show a greater knowledge and much better acquisitions, quantified by self-evaluation of the students and their exam results.

International audience; In the 21st century, luthiers are facing more and more international competition from the best factory-made and low cost hand-made instruments. Success depends on understanding what potential clients want, what is good or bad and avoiding poor decisions at any stage during making. Acoustics is useful in the quest to understand how instruments work as sound radiating machines. Knowing about vibration and sound radiation can inform about what is most important and what a particular change can be expected to do to the sound and playing feel. Since no two pieces of wood are identical luthiers need to adjust their geometry as work progresses to make use of their materials to best advantage. Acoustic measurements can thus be very helpful for such adaptive making as luthiers will have a lot more information on which to base their judgement and they are likely to be in a better position to decide what physical change is needed. In addition to learning about the vibro- acoustical function of instrument, it benefits them to learn about human hearing. It is extremely informative to take part in controlled listening experiments (though often frustrating and humbling). In order to help luthiers develop these skills in acoustics and critical listening, a training school was organised in 2017 at ITEMM (thanks to the support of the WoodMusick COST action). It has now been implemented as a yearly event, travelling across Europe to increase the possibility of exposure for both students and professionals at an affordable cost. Thus, the second training school took place in Belgium (Instrument making ? School of Arts, Ghent), the third one in Germany (Musical instrument making school, Mittenwald) and the next one wil happen in Italy (Politecnica delle Marche University, Ancona). It offers fundamental courses in (musical) acoustics along with critical listening tests and practical projects combining vibro- acoustical measurements and listening/playing tests.

International audience; The bow is an essential element for the sound production of bowed string instruments. The playing qualities of a bow, as evaluated by players, can be divided into two categories: those related to the playability (ability to execute technical gestures) and those related to the sound (influence of the bow on the sound of the instrument). Recent work (Ablitzer?s PhD, 2011) has allowed a better understanding of the influence of manufacture parameters on the mechanical behavior of a bow. However, links with playing qualities remain difficult to establish, as the interaction of the player and the bow has not been investigated yet. The production of the ?good? vibratory motion of a bowed string (Helmholtz motion) is mainly controlled by three parameters: the bow position on the string, the bow speed and the bow force. A procedure, based on optical motion capture and piezo-sensors on the bridge of two cellos (Wofford?s PhD, 2018), has been developed in order to measure these control parameters and explore whether and how players modify them when changing bows. In addition, the dynamics of the bow (modelled as a two degrees of freedom oscillator) has been added in a bowed string model to simulate the effect of the bow mechanical parameters on the bowed string vibratory motion. It shows a clear effect on the transients, which can then be compared with measurements made with piezo-sensors on the bridge, under the strings. Results of a preliminary study with one musician, one cello and two bows will be presented.