A novel subjective evaluation method for the sound quality of high-speed train compartments

Abstract

In this paper, a sound quality subjective evaluation method for the high-speed train compartments using semantic scoring method with equal even intervals is proposed. The acoustic metrics in the compartment was studied at different positions and under different operating conditions, and two subjective evaluation experiments using selected sound samples were carried out. The least square method was used to point-fit the sound pressure level values of the original signal, so as to select the appropriate short-term intercepting samples. To avoid the unclear distinction between samples evaluated with the unit interval scoring table, a novel evaluation method that equals even interval semantic scoring method was designed by increasing the interval between adjacent scoring numbers in the scoring table. The results show that the proposed method distinguishes the perceived auditory differences between samples more clearly compared with the unit interval semantic scoring table. And the correlation of the evaluation results obtained by using the equal even interval scoring method between participants is also higher, so the proposed method is suitable for evaluating the sound quality of the interior high-speed train compartments. Psychoacoustic attribute metrics such as loudness, roughness, and sharpness were calculated for the various samples, and loudness was found to have a strong correlation with the subjective evaluation results obtained using the proposed equal even interval method. This further verifies that, compared with unit interval scoring, the proposed method facilitates the sound quality evaluation of the interior high-speed train compartments.