Estimation of saxophone reed parameters during playing

Area of Science:

• Acoustics and Musical Instrument Physics
• Vibrational Dynamics and Control

Background:

• Accurate modeling of single reed instruments like the saxophone is crucial for understanding their acoustic behavior.
• Estimating physical parameters of the reed during performance is challenging due to complex interactions and dynamic variations.

Purpose of the Study:

• To develop and validate an iterative method for estimating single reed parameters in real-time during playing.
• To investigate the influence of different physical models on parameter estimation accuracy.
• To analyze sensor uncertainties and method limitations for a tenor saxophone reed.

Main Methods:

• Utilized an instrumented mouthpiece with sensors to capture reed tip movement during performance.
• Employed an iterative estimation approach to determine reed parameters based on measured data.
• Tested various physical models of reed tip oscillation, including non-linear stiffness and damping.

Main Results:

• The estimation method achieved small errors for low and medium dynamic levels (piano and mezzoforte).
• Sensor uncertainties and inherent method limitations were quantified.
• Model complexity required for accurate reed behavior description increases with dynamic levels.

Conclusions:

• The developed iterative method is effective for estimating single reed parameters at lower sound intensities.
• A non-linear oscillator model with stiffness and damping, but negligible mass, is most suitable for medium dynamics.
• Further research into more complex models is needed for high dynamic levels.