Figure 4.
The two time-frequency spectrograms are labelled a and b. Panel a presents a speech spectrogram with visible harmonic bands and vertical structures. Two rectangular highlighted regions mark areas of interest, including mid-frequency and higher energy segments. Panel b shows a corresponding spectrogram with a similar overall structure. The highlighted regions display clearer harmonic components and more distinct vertical patterns compared with panel a. Both panels show strong low-frequency energy bands along the bottom and time progressing horizontally across the image.Comparison of audio spectrum of sampling without SQG: (a) EN-AZS noise model result; and (b) EN-AZS speech model result