Hakanpää Tua, Waaramaa Teija, Laukkanen Anne-Maria
Speech and Voice Research Laboratory, Faculty of Social Sciences, Tampere University, Tampere, Finland.
Speech and Voice Research Laboratory, Faculty of Social Sciences, Tampere University, Tampere, Finland; Communication Sciences, University of Vaasa, Vaasa, Finland.
J Voice. 2023 Mar;37(2):293.e7-293.e23. doi: 10.1016/j.jvoice.2020.12.032. Epub 2021 Jan 23.
This study examines the effects of including acoustic research-based elements of the vocal expression of emotions in the singing lessons of acting students during a seven-week teaching period. This information may be useful in improving the training of interpretation in singing.
Experimental comparative study.
Six acting students participated in seven weeks of extra training concerning voice quality in the expression of emotions in singing. Song samples were recorded before and after the training. A control group of six acting students were recorded twice within a seven-week period, during which they participated in ordinary training. All participants sang on the vowel [a:] and on a longer phrase expressing anger, sadness, joy, tenderness, and neutral states. The vowel and phrase samples were evaluated by 34 listeners for the perceived emotion. Additionally, the vowel samples were analyzed for formant frequencies (F1-F4), sound pressure level (SPL), spectral structure (Alpha ratio = SPL 1500-5000 Hz - SPL 50-1500 Hz), harmonic-to-noise ratio (HNR), and perturbation (jitter, shimmer).
The number of correctly perceived expressions improved in the test group's vowel samples, while no significant change was observed in the control group. The overall recognition was higher for the phrases than for the vowel samples. Of the acoustic parameters, F1 and SPL significantly differentiated emotions in both groups, and HNR specifically differentiated emotions in the test group. The Alpha ratio was found to statistically significantly differentiate emotion expression after training.
The expression of emotion in the singing voice improved after seven weeks of voice quality training. The F1, SPL, Alpha ratio, and HNR differentiated emotional expression. The variation in acoustic parameters became wider after training. Similar changes were not observed after seven weeks of ordinary voice training.
本研究考察在为期七周的教学期间，将基于声学研究的情感声乐表达元素纳入表演专业学生声乐课程的效果。这些信息可能有助于改进演唱中的诠释训练。
实验性对比研究。
六名表演专业学生参加了为期七周的关于演唱中情感表达的音质额外训练。在训练前后录制歌曲样本。一个由六名表演专业学生组成的对照组在七周内录制两次，在此期间他们参加常规训练。所有参与者演唱元音[a:]以及一个表达愤怒、悲伤、喜悦、温柔和中性状态的较长乐句。34名听众对元音和乐句样本的情感感知进行评估。此外，对元音样本进行共振峰频率（F1 - F4）、声压级（SPL）、频谱结构（阿尔法比率 = 1500 - 5000 Hz的声压级 - 50 - 1500 Hz的声压级）、谐波与噪声比（HNR）以及微扰（抖动、闪烁）分析。
测试组元音样本中正确感知的表达数量有所增加，而对照组未观察到显著变化。乐句的总体识别率高于元音样本。在声学参数中，F1和声压级在两组中均能显著区分情感，而HNR在测试组中能特别区分情感。发现阿尔法比率在训练后能在统计学上显著区分情感表达。
经过七周的音质训练后，歌声中的情感表达得到改善。F1、声压级、阿尔法比率和HNR能区分情感表达。训练后声学参数的变化范围变宽。经过七周的常规嗓音训练后未观察到类似变化。
