For musicians, a beautiful note has appropriate pitch, loudness and timbre, and is elegantly articulated. On clarinet and saxophone, frequency, sound level, spectral envelope and articulation are controlled by breath, lip, tongue, vocal tract and fingers. Fingering effects are well understood and analysed. However, in spite of their importance to players, the effects of these other parameters on the sound and how players control them are neither discussed explicitly in pedagogical literature nor well studied in music acoustics. This thesis investigates these questions and the interesting physics behind them. One study was 'ecological': expert players played notes of different fundamental frequency, sound level, spectrum and articulation. While they played, the sound, the acoustical impedance spectrum of the vocal tract, the blowing pressure and its time variation, the tongue-reed contact and the coordination of these parameters were measured. In complementary experiments, the effects of the different control parameters were investigated in detail by varying them one at a time using a clarinet-playing machine. Different articulations (e.g. accent, sforzando, etc.) have different initial transients, with different rates of exponential increase in amplitude. Players vary these rates by increasing the blowing pressure P at different rates and coordinating this with the tongue release from the reed. Notes are terminated either by decreasing P below a threshold or by tongue contact: both produce exponential decreases in sound pressure. The exponential decrease rates are similar to those calculated from the bandwidths of the bore resonances. Quiet notes can be initiated by tonguing in the hysteresis region, where notes cannot start by increasing P only. Rapid tongue release from the reed and large initial tongue force produce substantial discontinuities in air flow and thus strong transients. Experienced saxophone players can vary the sound spectrum at constant pitch and loudness by changing the acoustic impedance spectrum of their vocal tracts, though the variation is smaller in the radiated sound spectrum than in the player's mouth. Harmonics of the radiated sound falling near peaks in the tract impedance are enhanced when the magnitude of the tract impedance peak is comparable with that of the instrument bore.