A multi-resolution sinusoidal model using adaptive analysis frame

The sinusoidal model has been applied to speech and audio signal coding, analysis/synthesis, modification, and various other fields. However, one drawback of this model is that the analysis frame size is generally fixed in analysing the speech and audio signal. As a result, optimal spectral resolution cannot be guaranteed to each sinusoidal component. In this paper, in order to overcome this drawback and to estimate the sinusoidal components more accurately, we propose a multi-resolution sinusoidal model using an adaptive analysis frame size. In the proposed scheme, the analysis frame size is varied based on the coarse and refined pitch characteristics. Experimental results have shown that the proposed model can achieve better performance than that of classical and multi-resolution sinusoidal models using a filter bank.