In real-life problems, various sensors are gaining serious attention due to their practical benefits in structural health monitoring. Typically, those sensors measures the time-series data which contained noise including other errors. The aforementioned issues make monitoring task complicated as often researchers struggle to access the measured sensors data. Hence, it is essential to select a practical and optimal approach to evaluate the data in order to extract useful features e.g. frequencies. Conventionally, the fast Fourier transformation (FFT) is widely used to estimate frequencies of any time-series due to its simplicity. However, many have reported that the aforementioned method may not yield good results when the measured signals are accompanied with significant level of noise or if the signals are non-stationary. To tackle the early discussed issue, some advanced methods can be utilised, for instance, the Wavelet Transforms (WT), the short-time Fourier transform (STFT), the Power Spectral Density (PSD) estimation, Spectrogram. In this study, an intra-comparison is studied via the use of the early discussed methods to evaluate their performances with the FFT. The outcome shows advanced methods (e.g. WT, STFT, PSD, Spectrogram) with better performance than its counterpart e.g. FFT. The results show that it is important to select an advanced time-Fourier analysis approach to obtain reliable results from the measured time-series data.