Real Time Spectral Analysis

Abstract

For signal analysis, Fourier transform and linear prediction methods are the main approaches. In recent years, wavelet transform which gives the expression of time-frequency domain has been receiving attention. In the wavelet transform, scale and shift operations are used to analyze regional waves in signal. In this paper, we propose a new method for signal analysis. Instead of the wavelet, short sine waves and short cosine waves which have slightly different frequencies and phases established beforehand are used. These frequencies and phases play the same role as a wavelet to analyze a regional wave in signal. However, our proposed method is different from the wavelet transform because it does not use multi-resolution analysis. In conventional spectral analysis methods, all data are prepared beforehand and calculations are carried out using these data. In our proposed method, spectrum is updated whenever the new signal data comes in. We show that the amount of calculation takes a remarkable reduction and it realizes real time spectral analysis. Finally, we show the vocal sounds analysis using the proposed method.