This paper presents experimental results that quantify the range of influence of vocal tract manipulations used in saxophone performance. The experiments utilized a measurement system that provides a relative comparison of the upstream windway and downstream air column impedances under normal playing conditions, allowing researchers and players to investigate the effect of vocal-tract manipulations in real time. Playing experiments explored vocal-tract influence over the full range of the saxophone, as well as when performing special effects such as pitch bending, multiphonics, and “bugling.” The results show that, under certain conditions, players can create an upstream windway resonance that is strong enough to override the downstream system in controlling reed vibrations. This can occur when the downstream air column provides only weak support of a given note or effect, especially for notes with fundamental frequencies an octave below the air column cutoff frequency and higher. Vocal-tract influence is clearly demonstrated when pitch bending notes high in the traditional range of the altosaxophone and when playing in the saxophone’s extended register. Subtle timbre variations via tongue position changes are possible for most notes in the saxophone’s traditional range and can affect spectral content from at least 800 – 2000 Hz .

Most single-reed woodwind instrument models rely on a quasistationary approximation to describe the relationship between the volume flow and the pressure difference across the reed channel. Semiempirical models based on the quasistationary approximation are very useful in explaining the fundamental characteristics of this family of instruments such as self-sustained oscillations and threshold of blowing pressure. However, they fail at explaining more complex phenomena associated with the fluid-structure interaction during dynamic flow regimes, such as the transient and steady-state behavior of the system as a function of the mouthpiece geometry. Previous studies have discussed the accuracy of the quasistationary approximation but the amount of literature on the subject is sparse, mainly due to the difficulties involved in the measurement of dynamic flows in channels with an oscillating reed. In this paper, a numerical technique based on the lattice Boltzmann method and a finite difference scheme is proposed in order to investigate the characteristics of fully coupled fluid-structure interaction in single-reed mouthpieces with different channel configurations. Results obtained for a stationary simulation with a static reed agree very well with those predicted by the literature based on the quasistationary approximation. However, simulations carried out for a dynamic regime with an oscillating reed show that the phenomenon associated with flow detachment and reattachment diverges considerably from the theoretical assumptions. Furthermore, in the case of long reed channels, the results obtained for the vena contracta factor are in significant disagreement with those predicted by theory. For short channels, the assumption of constant vena contracta was found to be valid for only 40% of the duty cycle.

Estimating the magnitude of the pressure reflection coefficient |R| and the end correction l at the open end of ducts is a critical procedure when designing or predicting the acoustic behavior of acoustical systems, such as exhausts, tailpipes, mufflers, loudspeaker enclosures and so on. For cylindrical ducts and plane waves, exact intricate solutions exist for two distinct open-end boundary conditions, namely for a thin-walled unflanged pipe and for a pipe terminated by an infinite flange. This work provides simple approximate expressions for |R| and l of cylindrical pipes terminated by circular flanges with finite radii. The expressions are obtained from a polynomial fit performed over the numerical results provided by a Boundary Element model, and is valid for Helmholtz numbers in the range 0 < ka < 3.0, as well as for 0 < a/b < 1, where a and b are the pipe and flange radii, respectively. When compared with the exact solutions for both the unflanged and the infinite-flanged pipe, the approximate formulae provide a maximum error of ~2% at the upper frequency limit (ka →3.0)

In this paper, the measurement of the absorption coefficient of non-locally reactive sample layers of thickness d1 backed by a rigid wall is investigated. The investigation is carried out with the aid of real and theoretical experiments, which assume a monopole sound source radiating sound above an infinite non-locally reactive layer. A literature search revealed that the number of papers devoted to this matter is rather limited in comparison to those which address the measurement of locally reactive samples. Furthermore, the majority of papers published describe the use of two or more microphones whereas this paper focuses on the measurement with the pressure-particle velocity sensor (PU technique). For these reasons, the assumption that the sample is locally reactive is initially explored, so that the associated measurement errors can be quantified. Measurements in the impedance tube and in a semi-anechoic room are presented to validate the theoretical experiment. For samples with a high non-local reaction behavior, for which the measurement errors tend to be high, two different algorithms are proposed in order to minimize the associated errors.