Reverberation Time in Traditional Courtyard Yue Opera Theatres

Abstract

The development of traditional Chinese theatres is closely related to the development of traditional Chinese opera. The acoustic quality of the theatre plays a crucial role in the performance effect of Chinese opera. As one of the five major traditional Chinese opera genres, Yue opera has a significant number of existing traditional theatres in the Shaoxing area. This study conducted on-site measurements of the reverberation time in traditional courtyard Yue opera theatres. The results indicate that the mid-frequency EDT values in the main viewing areas range from 0.6 to 0.7 s, and $T_{30}$ is around 0.7 s. Furthermore, aiming to investigate the optimal reverberation time for Yue opera performance spaces and the factors influencing preferences, subjective experiments using auralization technology and a paired comparison method were conducted. The results show that the optimal reverberation time for Yue opera performance spaces is around 0.8 s. There was no difference in the optimal reverberation time between the ordinary group and the experienced group, but the ordinary group tended to show a higher preference for shorter reverberation times. Compared to singing clips, recitation clips obtained higher preference scores for short reverberation times, indicating that audiences require higher clarity when listening to recitations.

1. Introduction

Traditional Chinese opera boasts a long history of evolution, developing numerous genres influenced by various regional cultures [1]. Among these, Yue opera stands out as one of the most widespread dialectal operas, often referred to as the “second national opera” or “Chinese Opera” internationally [2]. Its art form is comprehensive, integrating music, dance, fine art, and acrobatics. Renowned for its gentle and melodious tune, Yue opera excels in expressing emotions through singing. Yue opera originated in the mid-19th century in the vicinity of Sheng County, Shaoxing City. Its primary language is based on the Wu dialect, with some parts adopting a pronunciation similar to Mandarin Chinese during development [3]. Unlike the vocal range classifications in Western drama, Yue opera categorizes vocal styles based on role types, named “hangdang” [4]. This unique performance system features the stylized nature of character portrayal. Yue opera has four main hangdangs, namely, Laosheng (elderly males), Xiaodan (young females), Xiaosheng (young males), and Xiaochou (clowns or comedic men), each characterized by unique vocal properties [5]. This paper selected two representative hangdangs with strongly contrasting characteristics for study. Laosheng features a powerful and vigorous vocal style with many high notes, while Xiaodan has a gentle and delicate voice with more use of falsetto. The performance of Yue opera involves actors’ singing and recitation, accompanied by string instruments such as erhu and yangqin, as well as percussion instruments like drums and clappers [6].

The antecedent to Yue opera was story-singing, which combined folk tunes and Buddhist chants and narrated local tales on temporary stages amidst fields. The lyrics were mostly collected from conversations between working farmers. As the performance matured and gained popularity, it gradually went into traditional theatres in rural areas and towns [7]. With the flourishing development of Yue opera, its influence has continued to expand, and performances have become increasingly frequent in various areas. The thriving performances of Yue opera have also propelled the development and construction of traditional theatres. Zhejiang Province is one of the areas with the largest number of existing traditional theatres, of which Sheng County, as the birthplace of Yue opera, has a large number; almost every natural village has a temple and a theatre. Many theatres are still used for modern Yue opera performances and are important places for folk cultural activities.

Researchers have carried out studies on the acoustic characteristics of traditional Western theatres. Chourmouziadou and Kang [8] conducted sound-field simulations of ancient Greek and Roman theatres and pointed out that with the evolution of architectural materials and features, the acoustic characteristics of theatres changed greatly. Berardi et al. [9] conducted acoustic measurements and simulations of five ancient Roman theatres in Southern Italy, suggesting that these theatres exhibit low reverberation and loudness levels and require sound reinforcement systems when used for modern performances. Barkas [10] analysed the contribution of stage design to sound quality using 20 ancient Greek theatres as examples. Iannace et al. [11] researched the acoustics of a typology of a building erected in Ancient Greece and Ancient Rome named the Odeon and found that these buildings were well suited for music, songs, and speech. Sukaj et al. [12] discussed the acoustics of the Roman theatre of Benevento. The ERATO project evaluated the acoustic environment of existing ancient Greek and Roman theatres and proposed recommendations for restoration and improvement [13]. However, the ancient Greek and Roman theatres were primarily composed of stone and built on the mountain with a more open and wider space. Chinese traditional theatres, on the other hand, utilized wood as the primary building material and consisted of pavilion-style stages, courtyards, and corridors. There are significant differences in acoustic characteristics between traditional Chinese and Western theatres.

According to their architectural forms, traditional Chinese theatres can be categorized into open-air square theatres, courtyard theatres, and hall theatres. Among these, the courtyard theatre is the most significant and prevalent form. This study focuses on traditional courtyard Yue opera theatres. In recent years, there has been increased scholarly attention on traditional Chinese theatres. Most of the research on its acoustic characteristics is conducted from the perspective of objective evaluation by obtaining the objective acoustic parameters of the theatre through measurement or simulation. Wang [14] discussed the development, categories, and architectural features of traditional theatres in China and investigated the acoustic characteristics of traditional courtyard theatres in terms of reverberation, loudness, and clarity. Liu [15] conducted acoustic measurements and simulations of traditional Cantonese opera theatres, suggesting that the reverberation time in traditional theatres in the Lingnan area is short, and the strength index and clarity are high. Xue et al. [16] measured the reverberation time, early decay time, clarity, and strength index of eight traditional courtyard theatres in the Jiangnan area and analysed the acoustic characteristics of these theatres. Zhang et al. [17] conducted acoustic measurements on a traditional palace theatre in northern China, finding that it had good acoustic characteristics and was suitable for Peking opera performances. Moreover, the influence of architectural form changes on the sound field was examined through simulations. Current studies regarding the traditional Yue opera theatres in the Shaoxing area generally focus more on the architectural form, historical evolution, classification, preservation, and reconstruction. However, research on the acoustic characteristics of these theatres remains limited. In an era that lacked electroacoustic amplification systems, as the carrier of Yue opera art, the acoustic characteristics of traditional Yue opera theatres are crucial to the performance effect.

Different types of performing arts and music have varying demands for the acoustic properties of the performance space due to differences in the singing style, instrumental characteristics, and the aesthetic preferences of the audiences. Numerous studies have explored the preferred acoustic qualities of halls for diverse performing arts. Farina et al. [18] measured and compared the acoustic parameters of Italian opera houses and Japanese Noh theatres, finding that the frequency responses of the reverberation time and early decay time were completely different between the two types of theatres. Sarwono and Lam [19] investigated the subjective sound quality preferences for Indonesian traditional music, gamelan, revealing the audiences’ inclination toward clarity and intimacy. Yang et al. [20] conducted a study on the optimal reverberation time of Korean traditional music, gugak, and found that the optimal reverberation time was 1.7 s, while the room volume could affect the optimal value. Beranek [21] surveyed Western concert halls, reporting that the reverberation time in concert halls rated A+ and A ranged from 1.6 s to 2.1 s, while it was 1.3 s to 1.6 s in opera houses. For traditional Chinese opera performances, Zhu et al. [22] compared the evaluation dimensions of Western and traditional Chinese performance buildings and demonstrated that, due to cultural differences, the preferences for loudness and brightness were higher in the Chinese evaluation system. Wu et al. [23] conducted a subjective preference experiment on the reverberation time and sound pressure level of Chinese national music halls and found that the optimal reverberation time ranged from 0.8 s to 1.2 s. The subjects were satisfied with the spatiality and richness of the music when the reverberation time was 1.2 s. Wen et al. [24] studied the optimal values of reverberation time and the sound pressure level in Peking opera theatres and found that the optimal reverberation time was around 0.6 s. The younger generation tended to prefer longer reverberation times. Overall, the optimal objective acoustic parameters for Yue opera performance spaces require further research.

Reverberation time is the most important parameter significantly influencing the sound quality of theatres. It is essential to strike a balance between the clarity of recitation and the richness of singing and accompaniment. This study conducted on-site measurements of reverberation time in traditional courtyard Yue opera theatres. Moreover, subjective preference experiments were conducted, aiming to obtain the optimal value of reverberation time suitable for Yue opera performances.

2. Materials and Methods

2.1. On-Site Measurements

The traditional courtyard Yue opera theatres are semi-open spaces characterized by an open roof and enclosed sides, typically comprising a front hall, courtyard, balcony, main hall, corridors, and pavilion stage, as illustrated in Figure 1. It is a significant type of traditional Chinese architecture, primarily characterized by a wooden framework and bucket arch as its main structural elements. Wood is the predominant building material, and the theatre is enclosed by brick walls on all sides. The stage is positioned along the north–south axis of the courtyard and directly faces the main hall, extending outward on three sides to enhance interaction with the audience and facilitate better viewing experiences (Figure 2a). Behind the stage are dressing rooms for actors to change costumes and rest. The courtyard serves as the primary viewing area (Figure 2b), with symmetrically positioned balconies on both sides designated for women and children (Figure 2c). Furthermore, in some theatres, the stage also adopts a gallery connected to the main hall (Figure 2d), serving as a shelter from wind and rain and providing some reflection of sound from the top, and audiences can view performances underneath the gallery.

In this study, eight representative and typical traditional courtyard Yue opera theatres in the Shaoxing area were selected for acoustic measurements. The measurements were conducted in accordance with the ISO 3382-1 standard [25] and in an unoccupied condition with all doors and windows in the theatres closed. The measuring instruments included a dodecahedron sound source, a power amplifier, B&K 7841 Dirac architectural acoustics test and analysis software, a B&K ZE0948 USB sound card, a SENNHEISER MKH 800 microphone, a SENNHEISER SKP 2000 plug-on transmitter, and a SENNHEISER EM 2000 receiver. The instruments were calibrated in advance. The e-sweep signal was used as the sound source, which was arranged at two different positions on the stage, 1.5 m above the stage floor. As the layout of the theatre is symmetrical, measuring was only carried out on one side of the central axis. The receiving points were uniformly distributed in different viewing areas, such as the courtyard, corridors, balcony, and main hall, 1.2 m above the ground. No receiving points were set near the statues or worship areas in the main hall. Some of the corridors and balconies of the theatres have been converted into office spaces and so were not included in the measurements. Each receiving point was tested three times, and the average value was taken. Photographs of on-site measurements, theatre plans, and the arrangement of sound sources and receiving points are shown in

Table 1. Photographs of measured theatres and the arrangement of sound sources and receiving points.

Theatre	Photographs	Arrangement of Sound Sources and Receiving Points
Ma Taishou temple theatre
ShiWang temple theatre
Shi family ancestral hall theatre
Lu family ancestral hall theatre
DongGuo family ancestral hall theatre
Chenhou temple theatre
Song family primary ancestral hall theatre
Song family secondary ancestral hall theatre

.

2.2. Subjective Experiments

The subjective preference experiments were conducted in the laboratory using auralization technology [26]. Compared to the evaluation in the actual performance halls, the setup used allowed a quicker comparison of different halls’ sound quality, and objective parameters could be flexibly adjusted according to research needs [27]. The subjects listened to binaural signals of Yue opera with different reverberation times through headphone playback and made preference judgements.

Two representative Yue opera singing clips of Laosheng and Xiaodan were selected for this study. Apart from singing, recitation is also an important part of Yue opera performances, so a recitation clip was included. Specific details of the clips are displayed in

Table 2. Information on the clips used in the present study.

Clip Number	Clip Name	Category	Hangdang	Character	Duration
1	Wunü Baishou	singing	Laosheng	Yang Jikang	23 s
2	Lu You and Tang Wan	singing	Xiaodan	Tang Wan	38 s
3	Liang Shanbo and Zhu Yingtai	recitation	Xiaodan	Zhu Yingtai	28 s

. Yue opera performers were invited to perform the chosen clips with and without erhu accompaniment in the semi-anechoic chamber of the Department of Architecture, Zhejiang University, as shown in Figure 3. Dry signals were recorded using the NTi XL2 Portable Audio and Acoustic Analyzer, with a sampling frequency of 44.1 KHz and a quantization of 16 bits. Figure 4 illustrates the dry signals of clip 1. Before recording, the performers practised in the semi-anechoic chamber to familiarize themselves with the environment.

Referring to previous studies on Chinese national music [23,28] and Peking opera [24], five levels of mid-frequency reverberation times were chosen, which were 0.4 s, 0.8 s, 1.2 s, 1.8 s, and 2.5 s, and the frequency characteristics were kept flat. The binaural impulse responses were obtained through simulation using Odeon software (version 16.02). In accordance with Ando’s subjective preference theory [29], four independent objective acoustic parameters that describe the sound quality of a hall were used: the reverberation time ($T_{30}$), sound pressure level (SPL), initial time delay gap (ITDG), and interaural cross-correlation coefficient (IACC). After the reverberation time at the receiving point was adjusted to a set value, the ITDG and IACC were controlled within a just-noticeable difference (JND). Figure 5 shows the simulated binaural impulse response with a reverberation time of 0.8 s. The experimental signals were obtained by convolving the binaural impulse responses with the dry signals. The Sennheiser HD600 headphones were used to present the signals to the subjects, with the frequency response calibrated using its inverse filter. For each signal, the sound pressure level at the human ear position during headphone playback was measured using an artificial head (HMS IV.0, HEAD acoustics GmbH) and adjusted to 74.0 dBA.

The paired comparison method, which is widely applied in subjective psychological experiments [30,31,32], was used to obtain the preference score for each clip. There were 30 pairs of comparisons, consisting of 3 different Yue opera clips and 5 different reverberation times. All comparisons were played in a randomized order, and after each pair was played, the subjects were asked to make a preference judgement. The subjects took a 5-minute break for every 15 min of the experiment to avoid the impact of listening fatigue.

Exploring the preferences of different audiences for the sound quality of Yue opera can help maintain the sustainability of this traditional culture. Therefore, two groups of people with distinct backgrounds were selected to participate in the experiments. The first group, referred to as the ordinary group, consisted of students who lacked familiarity and experience with Yue opera. The second group, referred to as the experienced group, comprised members of a Yue opera club who had accumulated long-term experience attending live performances and were familiar with the selected clips. Each group consisted of 10 individuals with normal hearing abilities. Before the experiment, the participants were given a simple listening practice.

The coefficient of consistency was calculated to assess the circular errors that subjects made in the paired comparison test [33], and values lower than 0.8 were excluded from the analysis.

3. Results and Discussion

3.1. Results of Measurements

The measured results of mid-frequency reverberation time and clarity are presented in

Table 3. Measured reverberation times (average of 500 and 1000 Hz) and clarity (average of 500, 1000, and 2000 Hz).

Theatre		Reverberation Time (s) and Clarity (dB)
		Courtyard	Corridors	Balcony	Main Hall	Average
Ma Taishou temple theatre	EDT	0.65	0.61	0.67	0.83	0.69
	$T_{30}$	0.70	0.73	0.73	0.80	0.74
	$C_{80}$	7.62	7.13	6.30	4.48	6.38
Shi family ancestral hall theatre	EDT	0.59	0.65	0.61	0.71	0.64
	$T_{30}$	0.71	0.70	0.67	0.81	0.73
	$C_{80}$	7.66	7.70	7.65	6.10	7.28
ShiWang temple theatre	EDT	0.67	-	-	0.96	0.82
	$T_{30}$	0.84	-	-	0.91	0.88
	$C_{80}$	7.63	-	-	4.01	5.82
Lu family ancestral hall theatre	EDT	0.62	0.89	0.77	0.83	0.78
	$T_{30}$	0.74	0.79	0.71	0.79	0.76
	$C_{80}$	8.31	3.77	5.93	5.10	5.78
DongGuo family ancestral hall theatre	EDT	0.68	0.89	0.91	0.95	0.86
	$T_{30}$	0.90	0.94	0.92	0.95	0.93
	$C_{80}$	8.20	3.90	4.19	4.54	5.21
Song family primary ancestral hall theatre	EDT	0.58	0.78	-	0.76	0.71
	$T_{30}$	0.79	0.78	-	0.93	0.83
	$C_{80}$	7.25	4.76	-	3.96	5.32
Song family secondary ancestral hall theatre	EDT	0.56	0.83	0.67	0.80	0.72
	$T_{30}$	0.82	0.82	0.68	0.87	0.80
	$C_{80}$	6.99	2.30	4.39	3.89	4.39
Chenhou temple theatre	EDT	0.90	0.95	1.06	1.00	0.95
	$T_{30}$	1.03	1.05	1.05	1.09	1.06
	$C_{80}$	6.42	3.38	3.58	3.49	4.22

and Figure 6. The results indicate that the reverberation times in the courtyard, corridors, and balconies of the traditional courtyard Yue opera theatres are similar, with $T_{30}$ around 0.7 s and EDT ranging from 0.6 to 0.7 s, slightly shorter than $T_{30}$. The courtyard areas of the Shiwang temple theatre and the DongGuo ancestral hall theatre exhibit higher measured $T_{30}$ values, ranging from 0.8 to 0.9 s, which may be attributed to the recent extensive renovations and relatively new condition of these two theatres. Because it has the highest degree of enclosure, the main hall area exhibits richer sound reflections, with measured reverberation times ranging from 0.8 to 1.0 s, generally longer than other viewing areas. The stages of the Song family primary ancestral hall theatre, Song family secondary ancestral hall theatre, and Chenhou temple theatre are connected to the main hall through gallery structures, which led to sound reflections from the top, resulting in longer reverberation times. $T_{30}$ generally ranges from 0.9 to 1.0 s. The frequency characteristics of reverberation times in different viewing areas are relatively consistent due to the similarity in architectural materials and spatial layout in courtyard theatres, as is shown in Figure 6. The middle and high frequencies generally exhibit a smooth response, while the low frequencies show an increase of approximately 10% to 20%.

The clarity in the courtyards, corridors, and balconies of the theatres is generally high, with measured values ranging from 6 to 8 dB. The value in the main hall is lower, ranging from 3 to 6 dB. High clarity is a significant acoustic characteristic of traditional Yue opera theatre, contributing to enhanced music clarity during performances.

A comparison with the relevant measured results indicates that the mid-frequency reverberation times in traditional courtyard Yue opera theatres are longer than those in ancient Greek theatres (0.3 to 0.6 s) [13] and similar to those in other regions’ traditional courtyard theatres (0.7 to 0.8 s) [15]. Moreover, due to the lack of top reflections, they are generally shorter than those in traditional hall theatres (1.0 to 1.2 s) [34] and modern opera houses (1.4 to 1.5 s) [35]. Overall, traditional courtyard Yue opera theatres exhibit good intelligibility and clarity of music and speech, which can meet the audience’s requirements for clarity in lyrics and musical recitation. However, due to the relatively short reverberation time, the fullness of Yue opera music and accompaniment may be affected.

3.2. Results of Subjective Experiments

Figure 7a illustrates the mean preference scores of all clips at different reverberation times. The reverberation time of 0.8 s received the highest score, exhibiting a significant difference compared to the others. The subjects reported the most satisfaction with this reverberation time, which suggests that the optimal reverberation time for Yue opera halls might be about 0.8 s. The preferred order, with the score monotonically descending from the highest to the lowest, is 0.8 s, 1.2 s, 0.4 s, 1.8 s, and 2.5 s. The preference scores for clips with shorter reverberation times (0.4–1.2 s) were generally higher than those with longer reverberation times (1.8–2.5 s), which meant the audiences tended to prefer low reverberation. Notably, the subjects perceived long reverberation as a masking of the singing voice, which led to poor clarity of the lyrics. However, for clips with reverberation times as low as 0.4 s, the subjects reported that the erhu accompaniment sounded dry and harsh, resulting in distortion. Compared to the reverberation time in Western opera houses that Beranek recommended [21], Yue opera appears more suitable for theatres with shorter reverberation times. It is possible that the reverberation times in traditional Yue opera theatres generally ranged from 0.6 s to 0.8 s, which fostered the audiences’ listening habits and even the performance style over its long history. But, compared to another traditional Chinese opera, Peking opera, whose optimal reverberation time is about 0.6 s [24], Yue opera seems to have a longer optimal reverberation time. The reason might be that the melody of Yue opera, although developed from that of Peking opera, features stronger lyricism and a more relaxed tempo.

Figure 7b shows the mean preference scores for the ordinary group and the experienced group. Both groups have an optimal reverberation time of 0.8 s, while the ordinary group showed a higher preference score for a reverberation time of 0.4 s. Yue opera is based on the Sheng County dialect, and audiences who lack experience in attending live performances may be unfamiliar with the local language. Therefore, they would require a better understanding of the lyrics, leading to a higher demand for clarity.

Figure 7c shows the mean preference scores for different hangdangs. The clips of Laosheng received a significantly higher preference score than the clips of Xiaodan at a reverberation time of 0.4 s. This might be due to the fast tempo of Laosheng clips and the vocal style difference between hangdangs. For singing clips, a factorial ANOVA was performed for the factors of the subject group and hangdang. The results are given in

Table 4. The summary results of the ANOVA test on the preference scores of singing clips (significance levels: * = 0.05, ** = 0.01, *** = 0.001).

Reverberation Time	Group	Hangdang	Group × Hangdang
0.4 s	0.040 *	0.000 ***	0.809
0.8 s	0.438	0.619	0.149
1.2 s	0.759	0.483	0.172
1.8 s	0.536	0.464	0.278
2.5 s	0.107	0.000 ***	0.007 **

. The results show that at the two extreme reverberation time levels, i.e., 0.4 and 2.5 s, there is a statistically significant difference in the preferences of the different groups, while there is no significant difference in the preferences for the reverberation time values between 0.4 and 2.5 s. However, hangdang only had a significant difference in subjective preference at a reverberation time of 2.5 s.

The preference scores for singing clips and recitation clips of Xiaodan are displayed in Figure 7d. The recitation clips received a significantly higher preference score than the singing clips at a reverberation time of 0.4 s. Previous studies have shown that audiences prefer erhu clips with long reverberation, for which the optimal reverberation time is around 1.6 s [28]. Therefore, when listening to recitation without erhu accompaniment, audiences may have a higher requirement for clarity.

Compared to the optimal value, only the main hall areas in traditional courtyard Yue opera theatres provide a favourable acoustic environment, with a measured reverberation time close to 0.8 s. The values in other viewing areas are comparatively short. Theatres with gallery structures exhibit better acoustics, with the reverberation times in viewing areas approaching the optimal value. Considering the absorption of sound by the audience and orchestra in the occupied condition, the reverberation time in the theatre will be shorter than the measured values. Consequently, the theatre may lack early reflections during performances, and the fullness and liveliness of Yue opera music may be insufficient.

4. Conclusions

This study first carried out acoustic measurements of traditional courtyard Yue opera theatres in the Shaoxing area. Following these measurements, subjective preference experiments were conducted to investigate the optimal reverberation time in Yue opera performance spaces. The measurement results show that, in the unoccupied condition, the mid-frequency EDT values in the main viewing areas, such as the courtyard and balcony, of traditional courtyard Yue opera theatres range from 0.6 to 0.7 s, with a $T_{30}$ of around 0.7 s. The reverberation time in the main hall is longer than in other areas, ranging from 0.8 to 1.0 s. Additionally, theatres with gallery structures exhibit longer reverberation times, with measured mid-frequency values ranging from 0.9 to 1.0 s.

The results of the subjective preference experiments indicate that the optimal reverberation time for Yue opera halls is around 0.8 s. Compared to the experienced group, the ordinary group showed a higher preference for shorter reverberation time clips. The recitation clips obtained higher preference scores at a reverberation time of 0.4 s compared to the singing clips. Compared with the measured value, only the main hall areas in traditional courtyard Yue opera theatres are close to the optimal value. The theatres may lack early reflections during performances, and the fullness and liveliness of Yue opera music may be insufficient.

This study only investigated two different hangdangs in subjective experiments. In order to increase the generalizability of the results, future studies may consider incorporating more clips of different hangdangs. Moreover, this study only focused on the reverberation time; other acoustic parameters, such as the strength index and clarity, can also affect the sound quality of traditional Yue opera theatres. Further research is necessary to explore the measured and optimal values of these parameters and their potential interactions.