Research Report: Hybrid Clarinet Project

In the past thirty years we have seen a wide range of studies on the physical modeling of musical instruments using the waveguide technique. We now have at our disposal sophisticated models encompassing a large array of musical instruments. While the waveguide technique is both efficient and allows for the creation of effective models, it remains dependent on the precise control of the parameters of the model as well as on the quality of the excitation that is used to drive them. Many works on the control of waveguide physical models and on the modeling of nonlinear excitations have been carried out. However, the link between these two parameters is generally understudied. Indeed, for most musical instruments, the excitation is the element that has the greatest number of parameters to control and is thus the hardest element to model. The properties of the bore of a clarinet for example (but this also applies to most of the woodwind and brass instruments) can only be modified by tone holes, that are very discrete controllers (they can be opened or closed or half closed in some cases). On the other hand, the interactions between the mouthpiece and the player are extremely complex and difficult to simulate on a computer. This problem has been addressed in many ways in the past with different solutions for almost every case. Yamaha, for example, created a breath controller that worked with the VL1 synthesizer series1. For violins, Esteban Maestre developed a technique where the parameters of the physical model are controlled by gesture data acquired from real world performers[3]. The technique we present in this report uses a 3D printed mouthpiece combined with a system based on a piezo sensor and a loudspeaker to drive a simple physical model of a clarinet bore. The idea was to create a hybrid instrument based on a real mouthpiece and a virtual bore. We