Modern industrial process and household equipment more often use direct drives. According to European policy, Industry 4.0 and new Industry 5.0 need to undertake the effort required to ensure a sustainable, human-centric, and resilient European industry. One of the main problems of rotating machines is mechanical vibrations that can limit the lifetime of the final product or the machine in which they are applied. Therefore, analysis of vibration in electrical drives is crucial for appropriate maintenance of the machine. The present article undertakes an analysis of vibration measured at the laboratory stand with multiple dominant frequencies in the range 50–500 Hz. The fast Fourier transform (FFT) gives information about the frequency component without its time localisation. While the solution made available by the short-time Fourier transform (STFT) is able to overcome the problem of FFT, it still has limitations, particularly in terms of there being a lacuna in time and frequency localisation; accordingly, the need is felt for other methods that can give a good localisation in time and frequency. In the article, the continuous wavelet transform (CWT) was investigated, which requires selection of the wavelet function (kernel of transformation). The complex Morlet wavelet was selected with description of its central frequency and bandwidth. CWT and STFT time-frequency localisation capabilities were compared to investigate data registered from the direct-drive laboratory stand. CWT gives better frequency localisation than STFT even for the same frequency resolution. Vibration frequencies with near-locations were separated in CWT and STFT joined them into one wide pick. To ensure a good extraction of frequency features in electric drive systems, the author, based on analysing the results of the present study, recommends that CWT with complex Morlet wavelet be used instead of STFT.