Long Interpolation of Audio Signals Using Linear Prediction in Sinusoidal Modeling
Within the context of sinusoidal modeling, a new method for the interpolation of sinusoidal components is proposed. It is shown that autoregressive modeling of the amplitude and frequency parameters of these components allows us to interpolate missing audio data realistically, especially in the case of musical modulations such as vibrato or tremolo. The problem of phase discontinuity at the gap boundaries is also addressed. Finally, an original algorithm for the interpolation of a missing region of a whole set of sinusoids is presented. Objective and subjective tests show that the quality is improved significantly compared to common sinusoidal and temporal interpolation techniques of missing audio data.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is temporarily free for AES members.