The tonehole lattice cutoff frequency is a well-known feature of woodwind instruments. However, most analytic studies of the cutoff have focused on cylindrical instruments due to their relative geometric simplicity. Here, the tonehole lattice cutoff frequency of conical instruments such as the saxophone is studied analytically, using a generalization of the framework developed for cylindrical resonators. First, a definition of local cutoff of a conical tonehole lattice is derived and used to design “acoustically regular” resonators with determinate cutoff frequencies. The study is then expanded to an acoustically irregular lattice: a saxophone resonator, of known input impedance and geometry. Because the lattices of real instruments are acoustically irregular, different methods of analysis are developed. These methods, derived from either acoustic (input impedance) or geometric (tonehole geometry) measurements, are used to determine the tonehole lattice cutoff frequency of conical resonators. Each method provides a slightly different estimation of the tonehole lattice cutoff for each fingering, and the range of cutoffs across the first register is interpreted as the acoustic irregularity of the lattice. It is shown that, in contrast with many other woodwind instruments, the cutoff frequency of a saxophone decreases significantly from the high to low notes of the first register.

This paper presents how the bifurcation diagram of a saxophone model is affected by the contact force limiting the displacement of the reed when it strikes the mouthpiece lay. The reed impact is modeled by a nonlinear stiffness and damping activated by contact with the lay. The impact model is compared with the "ghost reed" simplification, where the reed moves through the lay unimpeded. Bifurcation diagrams in both cases are compared, in terms of amplitude of the oscillations and location of the bifurcations, on the solution branches corresponding to the first and second register. The ghost reed simplification has limited influence at low values of the blowing pressure parameter: the diagrams are similar. This is true even for "beating reed" regimes, in which the reed coincides with the lay. The most noticeable discrepancies occur near the extinction of the oscillations, at high blowing pressure.

The input impedance of woodwind instruments is characterized by at least two bands due to the lattice of open toneholes, a stop band at low frequencies, and a pass band at higher frequencies where the acoustic energy is able to propagate past the first open tonehole and into the lattice. The cutoff frequency that separates these two bands is an approximate value that is determined by the geometry of the lattice of open toneholes. It is expected that the frequency at which the stop band transitions to the pass band affects the sound produced by the instrument, but it is not known how this frequency affects the competition between self-sustained oscillation and radiation. A simplified model of a clarinet-like resonator is conceived such that the first input impedance peak and the cutoff frequency can be independently chosen. Experimental prototypes are built and the measured impedance of these prototypes is compared with the simulations. Resonators with very similar low frequency behavior but very different cutoff frequencies are then compared using digital synthesis to evaluate the influence of the cutoff frequency on sound production. The cutoff frequency impacts the synthesized pressure and acoustic volume velocity in the mouthpiece, particularly regarding the spectral content at high frequencies.

A saxophone mouthpiece fitted with sensors is used to observe the oscillation of a saxophone reed, as well as the internal acoustic pressure, allowing to identify qualitatively different oscillating regimes. In addition to the standard two-step regime, where the reed channel successively opens and closes once during an oscillation cycle, the experimental results show regimes featuring two closures of the reed channel per cycle, as well as inverted regimes, where the reed closure episode is longer than the open episode. These regimes are well-known on bowed string instruments and some were already described on the Uilleann pipes. A simple saxophone model using measured input impedance is studied with the harmonic balance method, and is shown to reproduce the same two-step regimes. The experiment shows qualitative agreement with the simulation: in both cases, the various regimes appear in the same order as the blowing pressure is increased. Similar results are obtained with other values of the reed opening control parameter, as well as another fingering.