Research Area

  • Title:
    Non Linear Modeling of Audio Systems

  • Keywords:

    Non linear audio systems, non linear modeling, distortion estimation

  • Description:

    System identification plays an important role in the field of digital audio systems. While linear systems can be fully characterized by their impulse responses in the time domain and their transfer functions in the frequency domain, non linear models have to be introduced for modelling most real-world devices. Considering this scenario, several efforts have been made to emulate and measure the acoustic behaviors of non linear electroacoustic devices employing digital signal processing techniques (dynamic convolution, Volterra series expansion and diagonal Volterra kernels, Hammerstein and Wiener models). In this field, the research is mainly focused on the study of non linear models working in real-time scenarios and in the measurement of non linear distortion with particular attention to harmonic and intermodulation distortion. In particular, the analysis is oriented towards solutions able to the online identification and distortion estimations of non linear systems, e.g., loudspeakers, hi-fi amplifiers, tube pre-amplifiers, guitar amplifiers and guitar pedals modeling.

  • Laboratory:

    Audio Digital Signal Processing Laboratory and Semi-Anechoic Chamber “E. Mattei” (Audio DSP Lab)

  • Contact Person:

    Stefania Cecchi

  • Collaborations:

    Università di Trieste (Italy), Università di Perugia (Italy)