Claude Cadoz

CORDIS-ANIMA: A Modeling and Simulation System for Sound and Image Synthesis: The General Formalism

CORDIS-ANIMA is a digital, real-time, object modeling and simulation system. The simulated objects can be seen, heard, and manipulated. This system has been built at the ACROE in several stages since 1978. It constitutes the nucleus of a creation tool for computer music (CORDIS) and animated images (ANIMA). The raison detre of this tool is to provide reproductions of physical objects from the real world. These objects can be musical instruments (in terms of objects that produce sound vibrations when subjected to some action), or they can be visible objects such as those we observe in the real world. nThe simulated copy (the simulacrum) of these objects is produced using a programmed digital device that can communicate with human operators through transducers associated with acoustic, visual, and gestural media. The gestural medium is a two- way channel; information can flow from the operator to the object (gestural action) and also from the object to the operator (tactile and kinesthetic feedback). nWhat we seek is total simulation. Such a simulation must restore not only the sensorimotor aspects of the object (the gestural, tactile, acoustic, and visual aspects) but also its structure and behavior. This is not merely a synthesis of the effects produced by the object but a synthesis of the intemal state of the object itself.

A Modular Feedback Keyboard Design

The modular feedback keyboard (MFK) design that we describe in this article was conceived and constructed at ACROE. It was first presented to the public on 17 October, 1989 in Grenoble at the opening celebration of our centers new studio. The system we present is the prototype of a more general device, and its construction is the result of research into the relationship between instrument and performer within the framework of a system for realtime sound synthesis and computer-assisted musical creation. The first feature of this device is that it is a gestural control system with feedback for the tactile senses-it enables actual fingering of a physical instrument to be simulated. The second important feature is that this is a modular system: the piano- or organ-type keyboard is its nominal form, but its mechanical morphology can be modified and easily configured for other applications or types of performance interfaces.

The CRM device: a force feedback gestural transducer to real-time computer animation

Abstract Movements in images or sounds are naturally produced by physical objects under the action of forces. These forces can be natural, such as those produced by gravity, wind or turbulence, artificial, as produced by manufactured motors, or human. Thus, to produce and control synthetic movements, computer simulation of physical objects can be monitored by specific interfaces, which provide the simulation of force generators, or convey real human gestural intervents. In this paper, we focus on the last approach. An analysis of human gestures driving a physical object is presented together with a force feedback gestural transducer we have developed, the CRM or modular retroactive keyboard. Finally, we show the computer context and paradigm in which we have developed this device, called the Instrumental Communication Interface (ICI).

Adaptive networks for physical modeling

Abstract This paper presents an original link between neural networks theory and mechanical modeling networks. The problem is to find the parameters characterizing mechanical structures in order to reproduce given mechanical behaviors. Replacing “neural” units with mechanically based units and applying classical learning algorithms dedicated to supervised dynamic networks to these mechanical networks allows us to find the parameters for a physical model. Some new variants of real-time recurrent learning (RTRL) are also introduced, based on mechanical principles. The notion of interaction during learning is discussed at length and the results of tests are presented. Instead of the classical {machine learning system, environment} pair, we propose to study the {machine learning system, human operator, environment} triplet. Experiments have been carried out in simulated scenarios and some original experiments with a force-feedback device are also described.

Analysis techniques for physical modeling networks

Association pour la Cr6ation et la Recherche sur les Outils dExpression (ACROE), Communication Langagiere et Interaction Personne-Systhme (CLIPS), and Leibniz (Neural Networks Group) Institut dInformatique et de Math6matique Appliqu6es de Grenoble (IMAG) 46 avenue F1lix Viallet, 38031 Grenoble Cedex, France nszilas@in-visio.fr Claude.Cadoz@imag.fr Analysis Techniques for Physical Modeling Networks

Lutherie virtuelle et interaction instrumentale. Une approche physique de l'instrument de musique numérique

Cet article propose une reflexion autour du geste instrumental musical, de sa nature physique et de sa place dans le monde numerique. Nous abordons dans un premier temps la difference fondamentale du lien geste-son entre instruments acoustiques et instruments musicaux numeriques, le premier cas etant regi par un couplage physique dynamique et le second par le traitement et la transmission dinformations et de signaux. Nous presentons ensuite une approche permettant le couplage physique avec des instruments virtuels simules, sappuyant sur le formalisme de modelisation physique CORDIS-ANIMA et lutilisation de systemes a retour deffort. En particulier, nous developpons les problematiques de cohabitation et de couplage physique dans le cadre de reseaux physiques multifrequence, Pour terminer, nous presentons des exemples dinstruments virtuels crees de cette maniere et discutons de linteret de creer un tel lien physique a linstrument dans le cadre numerique.

Exploring the Role of Dynamic Audio-Haptic Coupling in Musical Gestures on Simulated Instruments

Recent developments at ACROE-ICA have yielded a first audio-haptic modeller-simulator for real time simulations of mass-interaction physical models. This paper discusses the conceptual and technological considerations for a haptic platform that supports dynamic coupling for audio-haptic interactions, in particular in the case of physical coupling with virtual musical instruments. We present work concerning system calibration and metrology, then demonstrate a model developed with this platform, exhibiting dynamic gesture-sound coupling. The quality of the haptic device’s dynamics along with the modularity of mass interaction physical modelling allow for subtle and highly precise interaction with simulated objects, enabling intricate musical gestures relying on high-bandwidth audio-haptic coupling.

Interfacing Digital Waveguide with CORDIS ANIMA networks

The aim of this article is mainly to offer a link between the Digital Waveguide and the CORDIS‐ANIMA physical modeling formalism. The first one is highly a modular lumped physical modeling and simulation system based on the mass‐interaction paradigm while the second one offers accurate and efficient discrete time distributed models synthesized typically by delay lines and scattering junctions, in combination with digital filters. Both of them are widely developed and used in the domain of computer music field by scientists and artists. Although Digital Waveguide models have already been combined with Wave Digital Filters, they have never been exploited and integrated with CORDIS ANIMA networks. Wave Digital Filters are lumped models which are based on a scattering theoretic formulation which simplifies interfacing to Digital Waveguide models in contrast with the CORDIS ANIMA models. This research investigates the similarities of those formalisms, as well as focuses on the advantages of each modeling techn...

Analyse de modèles physiques, modèles physiques d’analyse. Synthèse sonore et analyse fréquentielle à l’aide des réseaux masses-interactions CORDIS-ANIMA

Notre propos se concentre sur les methodes danalyse et de modelisation pour lanalyse dediees aux reseaux physiques de type masse-interaction ou aux signaux produits par ces derniers. Dans le contexte de la synthese sonore pour la creation artistique se basant sur ce type de modelisation, de telles methodes saverent essentielles afin de soutenir la demarche des utilisateurs. Une premiere partie de cet article est dabord consacree aux proprietes de loscillateur harmonique tel quil peut etre construit avec le formalisme CORDIS-ANIMA. Puis, nous basant sur cet element simple, mais bien connu, nous presentons deux principes de calcul et danalyse des proprietes du signal que produira un modele cible, quil soit lineaire, non lineaire, simule ou non, en temps differe ou en temps-reel. Nous presentons enfin leurs implementations et utilisations au sein de lenvironnement GENESIS.

Sound synthesis and musical composition with the physical modeling formalism Cordis-Anima

The two historical starting points of Computer Music, in 1957, were digital sound synthesis, founded by Max Mathews (Bell Labs), and automatic composition, founded by Lejaren Hiller and Leonard Isaacson (University of Illinois). Digital sound synthesis and computer aided musical composition then developed and are today essential components of computer music. As said by Jean-Claude Risset, sound synthesis allowed to compose the sound itself. At the end of the 1970s, a new paradigm appeared, no more based on the synthesis of the sound signal, but on the simulation of the physical objects that produce the sound. The Cordis-anima language, created by Claude Cadoz and his colleagues Annie Luciani and Jean-Loup Florens at the ACROE lab, in Grenoble (France), is one of the most important representatives of this approach. We will present this language, which allows simulating the various components of a musical instrument, but also complex orchestras and dynamic objects with macro temporal behaviors. We will then...