This paper employs traditional acoustic features such as fundamental frequency, frequency perturbation, amplitude perturbation, Mel-frequency cepstral coefficients (MFCCs), and linear predictive cepstral coefficients (LPCCs) to study the identification of vocal characteristics in modern pop singing. By combining dynamic vocal fold image sequences, the vocal fold vibration components are extracted. Through the selection of key points on the vocal folds and the extraction of feature parameters, the dynamic analysis of vocal fold vibration characteristics under different singing techniques is conducted. There are significant differences in vocal fold spectra under different singing techniques. For example, in the growling singing state, there are more sound impurities and poorer tone purity, while in the normal chest voice technique, overtones are strong and account for a larger proportion. The glottal area change index under the normal chest voice technique ranges from 0.379 to 0.437, significantly lower than that of the growling and breathy voice techniques, but higher than that of the uniformly distributed vibration state. Additionally, the average vibration rate at three key vocal fold points is 3500–4500, with an amplitude of 5–11, showing significant differences from growling, localized vibration, and breathy voice techniques. The vocal fold vibration characteristics of modern pop vocal techniques exhibit high diversity and individuality, pointing to new research directions in music science.