Factors Influencing Natural and Cultural Soundscape Interactions on Perceptual Experiences in Forested–Historical Interface Areas

Abstract

The quality of the soundscape in historical districts is receiving increasing attention from urban governments due to its significant potential to highlight historical characteristics and enhance the acoustic environment of urban areas. However, there is still a lack of research on the relationship between natural and cultural soundscapes as they interact in historic areas. Using the historical area of Wuhou Shrine Museum in Chengdu as a case study, this study analyzed the differences in sound levels, sound source recognition, and subjective perception between two distinct spatial types: the historical street and adjacent urban forest. Additionally, structural equation modeling (SEM) was employed to explore the impact of sound source recognition and sound levels on subjective perception. The results reveal the following: (1) The soundscape interaction between the historical street and the adjacent urban forest exhibits a conflicting relationship, with cultural and natural soundscapes struggling to coexist harmoniously. (2) Within the historical region, L₁₀ has the strongest effect on subjective evaluation, while L₉₀ has the weakest. (3) Quietness is not always positively correlated with comfort and pleasure, indicating that a tranquil environment does not necessarily enhance pleasantness. These findings provide differentiated soundscape optimization strategies tailored to historical areas.

1. Introduction

As cities transition to urban governance that prioritize quality-of-life improvements, historical cultural districts and natural green spaces—key public areas—fulfill vital historical, cultural, and social roles [1]. These areas reflect urban cultural memory and local identity [2], while offering access to and integration with natural environments. Their spatial environments combine architectural heritage, natural landscapes, and human activity, creating rich and multifaceted atmospheres. Within this context, acoustic environments have emerged as critical determinants of spatial quality, integrating cultural soundscapes (e.g., human activities, markets) and natural soundscapes (e.g., wind, birdcalls, flowing water) [2]. The interaction between these two soundscape types directly shapes visitors’ perceptual experiences [3], yet their interplay remains underexplored in historic areas.

Soundscape was defined as the “acoustic environment as perceived or experienced and/or understood by a person or people, in context” [4], highlighting the importance of sound perception. Urban soundscape research focuses on the variations of urban sounds across different temporal and spatial contexts and their effects on the public [5,6]. Soundscapes are viewed as resources [7], which can be optimized to enhance the quality of urban acoustic environments [8,9]. Current research topics include: studies on sound sources, such as the identification of urban sound sources [10] and the characteristics of specific sounds (like bird songs) [11]; studies on the temporal and spatial variations of soundscapes, such as daily changes in noise complaints [12,13,14]; studies on sound evaluation, including how personal factors influence sound preferences [15,16]; studies on indicator construction, focusing on determining soundscape descriptors [17] and physiological survey indicators; studies on multisensory influences, such as the effects of audiovisual interactions on soundscapes [18], and studies on soundscape design, including the application of soundscape resources in landscape planning [19]. The above findings collectively demonstrate the effects of sound intensity and source on human perception, as well as differential impacts across distinct population groups. Additionally, such studies are integrated with each other. For instance, Francesco utilized a large cross-cultural sample to examine how sound source type (e.g., natural vs. artificial sounds) and intensity influence perceived pleasantness [20]. Separately, Chan investigated the role of individual characteristics and found that female participants reported lower acoustic comfort when exposed to nighttime sounds, sounds from transportation/built-up land uses, and mechanical/vehicular/aircraft noises [21]. Together, these studies provide a foundational framework for future research.

Urban forest, as vital landscape elements in urban areas, provide open spaces conducive to public visitation and positively influence citizens’ physical and mental health [22,23,24,25]. Soundscapes in historical landscapes are vital for promoting the authenticity of historical culture [26], enhancing urban historical perception [27], and contributing to urban historical preservation [28,29]. Existing soundscape research has made progress in two parallel directions relevant to this study: urban forest soundscapes and historical district soundscapes. However, there is still a lack of research on the interaction between these two directions.

On the one hand, research on the soundscapes of urban forests has gradually shifted from a relatively single focus on the restorative effects of natural sounds to exploring the interaction between artificial sounds and natural sounds [30,31]. Early studies, such as those by Ratcliffe et al. (2013), have shown that bird songs can enhance people’s attention recovery and reduce noise pollution caused by traffic [32,33]. And water sound, as an active sound source, also makes a significant contribution to enhancing tourists’ restorative effects [34,35,36]. In addition, some scholars have explored the potential connection between urban forest soundscapes and physiological benefits (such as reduced heart rate) [37]. Recent studies on human sounds in forests have shown that artificial sounds play a dual role in urban forests: low-intensity human activities (such as casual conversations) may enhance the perceived social vitality without significantly disrupting the sense of tranquility; while high-intensity artificial sounds may damage the overall environmental experience [38,39]. However, most of these studies focus on general urban parks or large woodlands, rather than forests adjacent to cultural heritage areas. This means that current research may not fully explore how forest soundscapes interact with heritage-specific human-made sounds (such as folk performances related to local history). Future research can further investigate this dynamic relationship [40].

On the other hand, research on the soundscapes of historical areas has also evolved from a single focus on the value of historical sounds to a comprehensive study of all elements of the soundscapes in historical areas [41]. For example, Zhang et al. (2016) analyzed the soundscape of Chinese Han Buddhist temples and found that religious sounds (such as chanting) enhanced tourists’ perception of cultural authenticity [42]; Kim et al. (2022) explored the bell sounds in the historical center of Daejeon, revealing how these sounds shape people’s collective “acoustic memory” of the site’s past [43]. Subsequently, researchers began to explore the interrelationship between cultural soundscapes and natural soundscapes within historical areas. For example, Germán et al. (2018) demonstrated that in the historically significant monumental sites within semi-natural environments, natural sounds enhance the overall sense of pleasure [44]. Bartalucci et al. (2020) systematically explored the application of soundscape research in the field of cultural heritage, emphasizing the symbiotic relationship between natural soundscapes and cultural soundscapes [45]. Đorđević et al. (2023) explored sacred soundscapes in Serbia, highlighting the need to extend archaeological acoustic studies of sacred sites to include surrounding outdoor soundscapes [46]. Jo et al. (2021) surveyed Cheong-ju, finding that both anthropogenic and natural sounds dominate the soundscape of historical areas [47]. These studies indicate that people have begun to pay attention to the interaction between nature and culture, where natural sounds serve as temporal anchors connecting the past and the present.

Despite these advancements, the integration of natural sounds and cultural sounds in historical districts remains insufficient. Moreover, most current studies focus on the relationship between cultural and natural sounds in historical areas, with little discussion of their spatial interactions. Historical areas typically encompass historic districts and abundant green spaces. Within urban soundscapes, the soundscapes of historic districts and green space are considered worthy of preservation [29]. Historic districts represent cultural soundscapes dominated by human activities (visitor conversations, tour guides, vendor calls), while green space emphasize natural soundscapes with biophony and geophysical elements (wind, water, bird songs). The interaction between these soundscapes significantly shapes visitor perception [48], yet limited research explores this interplay—especially in settings where historical culture and natural environments coexist. Previous studies often analyze one soundscape type in isolation, neglecting their coexistence and interaction in the same area. This gap hinders the development of integrated soundscape optimization strategies for historic–nature interface areas.

This study addresses this gap by investigating how natural (urban forest) and cultural (historic street) soundscapes interact to influence visitor perception in the Wuhou Shrine Museum area, Chengdu. Existing studies mostly analyze the soundscape of a single space based on general soundscape theories, but pay relatively little attention to the composite characteristics of the ‘historical–natural interface area’. This study extends the ‘expectation–perception matching theory’ [49,50] from general urban scenarios to heritage contexts by introducing the expectation mechanism. It confirms that subjective perception in historical scenarios not only depends on sound intensity/sound sources, but is also regulated by cultural expectations and audiovisual consistency, which will supplement the soundscape theory in historical–natural composite scenarios. Specifically, we examine: (1) the spatial differences in sound intensity, sound source recognition, and subjective experience between historic streets and adjacent urban forest; (2) the impact of sound source recognition (biophony, anthropophony) and sound intensity metrics (L₁₀, L₉₀, etc.) on subjective evaluations. By integrating objective acoustic measurements and subjective surveys, this research aims to provide a theoretical basis for developing tailored soundscape optimization strategies in historical areas.

2. Methodology

This study adopts a three-level framework of data collection–analysis–modeling to carry out the research (Figure 1): In the first stage, objective data of sound pressure levels at 20 research points were collected using a sound level meter (AWA 6628+, Hangzhou Aihua Instruments Co., Ltd., Hangzhou, China), and 113 pieces of subjective perception data were collected simultaneously through structured questionnaires. In the second stage, kernel density analysis (ArcGIS for Desktop 10.6) was used to analyze the spatial distribution characteristics of sound intensity and sound sources, and Spearman correlation analysis was used to explore variable correlations. In the third stage, a SEM was constructed based on the soundscape perception theory to verify the influence paths of sound intensity, sound source identification, and individual information on subjective perception. All analyses passed reliability and validity tests to ensure reliability.

2.1. Study Area

The Wuhou Shrine Museum, located in Wuhou District, Chengdu, Sichuan Province, China, is a national scenic area. The museum comprises three main sections: the Three Kingdoms Historical Relic Area (Cultural Relics Zone), the Western Area (Three Kingdoms Cultural Experience Zone), and the Jinli Folk Area (Jinli Ancient Street), covering a total area of 150,000 m². Jinli Ancient Street (550 m length) features vibrant cultural soundscapes including folk performances, vendor calls, and historical ambient sounds, while the Historical Relic and Western Areas offer contrasting natural soundscapes characterized by water features, forested areas, and bamboo groves acoustics [51]. In this study, Jinli Ancient Street is a historical street, while Historical Relic and Western Areas are an urban forest. At the same time, this forest has been developed into an urban park according to local conditions.

This study selected 20 survey locations within the Wuhou Shrine Museum for field research (Figure 2). These points were strategically placed in areas of high visitor activity and featured distinct soundscape characteristics. The survey points included intersections of pathways and green spaces, areas around water features, shaded resting zones, and temporary gathering areas. Specifically, Points 2, 5, and 6 were located within Jinli Ancient Street, Points 1, 3, 4, 7, and 8 were situated at the interface of Jinli Ancient Street and green spaces, and Points 9 to 20 were distributed across the museum’s green spaces. The spatial conditions of each measuring point are shown in Figure 3.

2.2. Theoretical Basis

This study was grounded in the evolutionary process of soundscape research, focusing on the interaction between natural and cultural soundscapes and their influence on perceptual experiences.

The concept of “soundscape” (defined by ISO 12913-1:2014 [4] as a perceived acoustic environment) emphasizes human perception as its core, evolving from early focus on natural soundscapes (biophony, geophony) [52] to integrating cultural soundscapes (anthropophony) [6]. As for humanistic soundscapes, especially those in historical districts, relevant studies usually focus on their core sound source composition (such as folk sounds) and explore how they shape perceptual experiences [3]. Recent studies usually involve the impact of different sound source types [10,11] and sound intensities [12,18,20] on human perception, and also explore the differences in perception among different groups [15,21]. As for humanistic soundscapes, especially those in historical blocks, relevant studies usually focus on their core sound source composition (such as folk sounds) and explore how they shape perceptual experiences [3].

Our study follows this process: we first identify natural (forest) and cultural (historic street) sound sources, measure sound intensity in different environments, then analyze their interaction and impact on subjective perceptions (e.g., comfort, harmony), aligning with the theoretical logic of soundscape research.

2.3. Content of Subjective and Objective Data Collection

Objective data were collected on-site using an AWA 6628+ Class I sound level meter, covering four indicators: L₁₀, L₅₀, L₉₀, and L_Aeq. Specifically, L₁₀ (peak sound pressure), L₅₀ (median sound pressure), and L₉₀ (background sound pressure) represent sound levels exceeded during 10%, 50%, and 90% of the survey time, respectively; the difference (L_10-90) reflects peak-to-background sound variability. L_Aeq denotes the equivalent continuous sound pressure level. Data collection was conducted between 8:00 and 10:00 AM (local time), as this period captures the peak of visitor activity in the study area in the morning. This includes folk performances, early sightseeing, and vendor operations, thereby reflecting typical soundscape interactions between the historical street and forest. Acoustic surveys were performed at 20 locations, with each survey lasting 1 min and repeated three times per location, which can avoid capturing non-representative events (e.g., a sudden loud noise). The Coefficient of Variation (CV) of the sound pressure level (SP) after collection is shown in

Table 1. Coefficient of Variation (CV) of sound pressure level data.

Survey Point	CV (LAEQ)	CV (L10)	CV (L50)	CV (L90)
1	72.44%	150.44%	98.91%	129.69%
2	72.19%	15.72%	51.66%	17.56%
3	625.88%	615.62%	369.98%	116.00%
4	295.89%	416.44%	199.39%	31.07%
5	313.68%	383.47%	145.47%	78.93%
6	296.90%	347.84%	140.22%	49.77%
7	1091.87%	361.54%	203.22%	227.53%
8	312.80%	579.90%	209.60%	91.55%
9	381.37%	59.03%	410.19%	232.47%
10	220.60%	470.37%	115.80%	108.19%
11	112.20%	466.86%	103.72%	10.81%
12	288.90%	395.40%	156.10%	144.52%
13	115.65%	57.23%	45.07%	18.80%
14	131.93%	71.03%	261.01%	85.10%
15	172.89%	172.55%	142.38%	120.89%
16	352.60%	28.99%	526.79%	41.49%
17	137.22%	151.50%	600.47%	190.92%
18	131.51%	317.87%	236.77%	735.20%
19	463.38%	184.71%	118.29%	152.34%
20	228.52%	340.23%	336.50%	213.59%

, which demonstrates good reliability. SPL metrics were prioritized to capture dynamic sound variations in historical areas, where human activities and cultural sounds dominate the acoustic environment.

Subjective perception data were gathered through a structured questionnaire consisting of three sections:

(1) Demographic information: includes variables such as gender (male, female), age group (≤18, 18–35, 36–45, 46–65, 65–80, >80), educational background (high school or below, undergraduate/diploma, graduate or above), place of residence (resident, non-local visitor), occupation (student, employed, retired, other), visit frequency (first-time visitor, 2–3 visits, weekly, multiple times per week, at least once per month), and monthly income (≤1500 CNY, 1500–3000 CNY, 3000–5000 CNY, 5000–10,000 CNY, >10,000 CNY).

(2) Sound source perception, where participants identified geophysical, biological, and anthropogenic sounds. To ensure consistency, participants received pre-survey training. Geophysical sounds refer to non-biological natural sounds such as wind, rain, and thunder. Biological sounds include those produced by living organisms, such as birdsong and human conversation, while anthropogenic sounds arise from human-made activities.

(3) Sound environment evaluation: including pleasure, comfort, quietness, and harmony. Pleasure refers to an individual’s perceived level of enjoyment and satisfaction within the soundscape. Comfort indicates a sense of relaxation and ease, while quietness reflects the perception of low-intensity sound levels in the environment. Lastly, harmony denotes the perceived balance and cohesion among various auditory elements. Soundscape evaluation used a 5-point Likert scale assessing pleasure, comfort, quietness, and harmony. Scoring criteria: very unpleasant/uncomfortable/quiet/harmonious (+1), somewhat unpleasant/uncomfortable/quiet/harmonious (+2), neutral (+3), somewhat pleasant/comfortable/quiet/harmonious (+4), very pleasant/comfortable/quiet/harmonious (+5).

2.4. Experimental Design

To ensure the feasibility and consistency of data collection, the formal survey was conducted on sunny weekdays in April 2021. The target population included two groups: (1) tourists aged 18–80 years visiting the Wuhou Shrine Museum (excluding those in a hurry or with hearing impairments, unique to each survey site); (2) 10 graduate student volunteers with prior soundscape research experience (participated in surveys at all 20 sites to control individual bias and ensure cross-site data consistency).

The overall experimental process includes: Questionnaire Validation and Pre-Training; Sampling Design; and Formal Survey Procedure.

(1)

Questionnaire Validation and Pre-Training:

Questionnaire Design Evaluation: We verified the validity of the questionnaire items through the following methods: expert review (3 soundscape scholars evaluated content validity); pre-testing (n = 30) to optimize wording and improve the accuracy of responses.

Pre-training: All participants received pre-training before the formal experiment to enhance the accuracy of the experiment. The training content was standardized as follows: (1) explaining the research purpose; (2) clearly defining biological sounds (such as bird songs, human conversations), geophysical sounds (such as wind, water flow), and anthropogenic sounds (such as vendors’ cries, broadcasts) with audio examples; (3) demonstrating the sound recording method (recording identified sounds every 20 s). Volunteers received additional advanced training, including a 5-min pilot test, to practice distinguishing subtle sound differences (such as the sounds of wind and water) and ensure consistency in recording standards. Tourists received a 5-min brief audio-visual guidance, focusing on sound category identification and subjective evaluation skills (using a 5-point Likert scale).

(2)

Sampling Design:

We have standardized the number of samples for the questionnaire survey and the three key sampling settings for tourists/volunteers as follows:

Quantity setting: The pre-test results (n = 30) showed that 3–5 respondents per site are sufficient to capture the spatial variation in sound perception. With 20 sites, the target sample size was 100–120, and we ultimately collected 113 valid responses (5–7 per site), which met the requirements for statistical and spatial representativeness.

Tourist sampling setting: At each of the 20 pre-determined survey points, tourists were selected using simple random intercept sampling. Investigators approached every 3rd tourist passing through the point (to avoid bias) and invited them to participate, ensuring a diverse range of demographics (age, origin). This method was chosen to reflect the natural variation of visitors in the area.

Volunteer sampling setting: The 10 volunteers were pre-selected based on their experience in soundscape research (to ensure consistency in sound recognition). They were assigned to all 20 survey points using a rotational schedule (each volunteer visited 4–5 points, which included parks and streets, randomized to avoid temporal bias). This design controlled for individual differences in perception, allowing valid cross-site comparisons.

(3)

Formal Survey Procedure:

Acoustic measurement: At each point, a sound level meter (AWA 6628+) recorded L₁₀, L₅₀, L₉₀, and L_Aeq for 1 min, repeated 3 times (at 10-min intervals) to capture short-term variations.

Sound perception test: Tourists: Closed eyes and listened for 3 min, recording identified sounds every 20 s (9 intervals) on a pre-designed form; Volunteers: Completed a 5-min test with the same 20-s interval (15 recordings) to capture more detailed sound dynamics, leveraging their training.

Questionnaire administration: Immediately after the perception test, participants rated pleasure, comfort, quietness, and harmony using a 5-point Likert scale. Responses were collected on-site to avoid memory bias.

Data synchronization: All records (acoustic data, perception logs, questionnaires) were labeled with the survey point ID, time, and participant ID for cross-referencing.

2.5. Reliability and Validity of Experimental Data

A total of 120 questionnaires were distributed in the experiment, and 113 valid responses were collected (N = [5, 7]). We conducted reliability and validity tests to assess the quality of the collected questionnaire data. The reliability test evaluates the internal consistency and accuracy of the quantitative data. Results show that the Cronbach’s alpha coefficient for subjective perception is 0.811, exceeding the acceptable threshold of 0.7, indicating high reliability. The Corrected Item-Total Correlation (CITC) values for pleasantness (0.672), comfort (0.771), quietness (0.703), and harmony (0.491) are all above 0.4, suggesting good item correlation and supporting the internal consistency of the scale. These results indicate that the scale is well-designed and suitable for further analysis.

The validity was assessed through the Kaiser–Meyer–Olkin (KMO) value and Bartlett’s sphericity test. The KMO value is 0.776 (>0.60), confirming sampling adequacy for factor analysis. Bartlett’s test of sphericity was significant (p < 0.001), indicating that the data are suitable for structure detection. Exploratory factor analysis extracted three common factors, consistent with the structure of the questionnaire. The cumulative variance explained after rotation is 74.832%, exceeding the 50% threshold, which demonstrates that the items effectively capture the underlying constructs of interest.

2.6. Experimental Data Analysis

The data analysis included the following three major components:

(1) Spatial Distribution Analysis: Kernel density analysis was used to visualize the spatial distribution of sound intensity and auditory perception, thereby identifying areas where soundscape interactions occur. Kernel density analysis uses the default bandwidth optimization method of ArcGIS 10.2 (Silverman’s rule of thumb) to calculate the bandwidth h = 0.9 × min (standard deviation, interquartile range/1.34) × n^(−1/5), where n = 20 (the number of research points), and the final bandwidth value is 50 m. This bandwidth not only avoids fragmented hotspots caused by excessively small bandwidth but also prevents excessive spatial smoothing due to excessively large bandwidth, which is in line with the spatial scale of the study area (150,000 m²).

(2) Correlation Analysis: Spearman’s rank correlation was used to explore associations between acoustic indicators and subjective responses; Kernel density analysis was used to explore the spatial distribution of variables.

(3) Causal Analysis: A SEM was developed to identify key factors influencing subjective perception.

Finally, the study examined synergy and conflict between street and forest soundscapes—that is, the interaction between cultural (street) and natural (forest) auditory environments within the historical district. For instance, commercial sounds from the street may complement biological sounds from adjacent green spaces (e.g., vibrant street chatter alongside birdsong), or they may conflict (e.g., vendor noise disrupting forest tranquility). The analysis also considered potential influencing factors of this interaction, including acoustic parameters (e.g., sound intensity, type) and visitor demographics (e.g., age, site familiarity).

2.7. Structural Equation Modeling (SEM)

SEM is a multivariate data analysis method used to explore the interrelationships among multiple latent variables. In this study, SEM will be employed to further analyze how physical sound intensity and subjective sound source recognition influence subjective sound evaluations. The SEM approach adopted in this study includes three components: data processing, Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

2.7.1. Data Processing

For conducting SEM analysis, nested data (individual data within 20 sites) were aggregated to the site level (1 row of data per site). This study employed a data aggregation method for data preprocessing, whose steps include:

(1) Group mean: Grouping by site number, the mean value of each indicator within the group was calculated. Among them, the objective acoustic indicators (L₁₀, L₅₀, L₉₀, L_Aeq) were the average of all measured values at each site; the subjective experience indicators (pleasure, comfort, quietness) were calculated using a weighted mean, where the weight was the proportion of the number of volunteers/visitors at the site to the total number of people at that site; the sound source identification and individual information indicators were calculated using the above-mentioned weighted mean method.

(2) Standard error (SE) calculation: For the aggregated mean of each site, the standard error was calculated to reflect the degree of data variation.

(3) Data aggregation: After aggregation, 20 rows of data were formed (1 row for each site), including the site mean and standard error of all variables, which were used for subsequent SEM analysis.

2.7.2. EFA and CFA

According to Spearman’s rho correlation analysis (

Table 2. Spearman’s rho correlation analysis of Sound Intensity, Sound source recognition, and Subjective Perceptions.

	LAeq (1)	L10 (2)	L50 (3)	L90 (4)	L10−90 (5)	Geophony (6)	Biophony (7)	Anthropophony (8)	pleasant Level (9)	Comfort Level (10)	Quietness (11)	Harmony (12)
1	1.000
2	0.908 **	1.000
3	0.678 **	0.747 **	1.000
4	0.592 **	0.532 *	0.887 **	1.000
5	0.672 **	0.795 **	0.346	0.051	1.000
6	−0.226	−0.214	−0.055	−0.108	−0.077	1.000
7	0.298	0.150	0.072	0.101	−0.074	−0.135	1.000
8	−0.086	−0.149	−0.123	−0.060	−0.280	−0.328	0.391	1.000
9	0.103	0.369	0.195	−0.094	0.365	0.351	−0.236	−0.331	1.000
10	0.280	0.509 *	0.495 *	0.196	0.505 *	0.351	−0.309	−0.371	0.803 **	1.000
11	0.135	0.388	0.348	0.085	0.446 *	0.299	−0.496 *	−0.412	0.777 **	0.909 **	1.000
12	−0.143	−0.107	0.048	−0.039	−0.185	0.473 *	−0.218	−0.146	0.384	0.469 *	0.507 *	1.000

Yellow background, * p < 0.05, Pink background, ** p < 0.01.

), the variables of geophysical sound and harmony have a relatively low correlation with other variables. In addition, during the preliminary principal component analysis, geophysical sound was classified into the component related to subjective feelings, which is inconsistent with the general expectation. Therefore, when conducting the EFA, the two variables of geophysical sound and harmony were removed.

Through EFA (

Table 3. EFA of SEM.

	Variable Name	Component				Explained Variance/%
		1	2	3	4
1	individual information-age				0.761	23.148
2	individual information-educational information				−0.808
3	individual information-place of origin				−0.79
4	individual information-occupation				0.698
5	individual information-visiting frequency				0.824
6	individual information-monthly income				0.603
7	Sound intensity-LAeq	0.884				22.672
8	Sound intensity-L10	0.913
9	Sound intensity-L50	0.937
10	Sound intensity-L90	0.886
11	subjective perception-pleasant level		0.886			15.951
12	subjective perception-comfort Level		0.878
13	subjective perception-quietness		0.799
14	sound source recognition-biophony			0.763		11.756
15	sound source recognition-anthropophony			0.803
Total Explained variance/%		73.527

Rotation method: Varimax.

), four common factors were extracted, with a total variance explained of 73.527% (greater than 50%, indicating good explanatory power), and the KMO test is 0.643 (greater than 0.6, meeting the analysis requirements). Common Factor 1 includes the variables “L_Aeq”, “L₁₀”, “L₅₀”, and “L₉₀”, which can be categorized as sound intensity, accounting for 22.672% of the variance. Common Factor 2 includes the variables “pleasure”, “comfort”, and “quietness”, which can be categorized as subjective perception, accounting for 15.951% of the variance. Common Factor 3 includes the variables “biological sound” and “anthropogenic sound”, which can be categorized as sound source recognition, accounting for 11.756% of the variance. Common Factor 4 includes variables such as “age”, “educational background”, “place of origin”, “occupation”, “visit frequency”, and “monthly income”, which can be categorized as individual information, accounting for 23.148% of the variance. The extracted factors align well with the predefined categories.

The common factors derived from the EFA were used as latent variables in the SEM, with their corresponding indicators treated as observed variables. To further verify the robustness of the measurement model, CFA was conducted, including both reliability and validity assessments. To adapt to the aggregated data structure of the final SEM, aggregated data at the measurement point level (n = 20) was used for CFA validation.

Reliability was evaluated using Cronbach’s alpha (CA) for each latent construct. As shown in

Table 4. Reliability and Validity Analysis of the SEM Measurement Model.

Latent Variable	Observed Variable	STD. ESTIMATE	C.R	AVE	Cronbach’s α
Sound intensity	LAeq	0.721	0.921	0.747	0.921
	L10	0.799
	L50	1
	L90	0.912
Sound source recognition	Biophony	1	0.858	0.757	0.835
	Anthropophony	0.716
Subjective perception	Pleasant level	0.882	0.945	0.852	0.886
	Comfort level	0.995
	Quietness	0.889
Individual information	Age	0.531	0.751	0.502	0.721
	Educational information	−0.691
	Place of origin	−0.62
	Occupation	0.928
	Visiting frequency	0.531
	Monthly income	0.476

, the CA values for “sound intensity”, “sound source recognition”, “subjective perception”, and “individual information” were all above 0.7, indicating strong internal consistency and reliability of the measurement model.

To assess convergent validity, Average Variance Extracted (AVE) and Composite Reliability (CR) were calculated. Convergent validity is considered acceptable when AVE > 0.5 and CR > 0.7. Based on the CFA results computed in AMOS Graphics 24 (Table 4), all three latent constructs showed AVE values exceeding 0.5 and CR values greater than 0.7, confirming that the measurement model demonstrated good convergent validity.

2.7.3. Structural Equation Model Construction and Hypotheses

Based on the theoretical framework, research objectives and previous studies (

Table 5. Theoretical Basis for SEM Path Settings.

Path	Core Theoretical Basis	Literature Support
Sound intensity --> Subjective experience. (Ha)	Sound intensity, as a physical stimulus, directly affects subjective arousal and emotional experience.	[6,18]
Sound source --> Subjective experience. (Hb)	Sound source types influence subjective experience through “emotional association”.	[11,15]
Individual information --> Subjective experience. (Hc)	Individual characteristics regulate subjective perception through “soundscape familiarity” and “cultural background”.	[10,18,20,24]
Sound intensity <--> Sound source (Hab)	There is a “dynamic synergistic effect” between sound intensity and sound sources; changes in the intensity of different types of sound sources directly affect subjective perception.	[11,20]
L10 <--> L90	Both belong to the sound intensity index system.	[4,17]
L50 <--> L90	Both belong to the sound intensity index system.	[4,17,53]
L50 <--> Subjective experience	Median sound intensity can enhance perceived controllability and directly affect comfort.	[3,54]
L90 <--> Subjective experience	Background sounds can reduce auditory fatigue and directly affect the evaluation of “sense of quietness”.	[29,55]
Visiting frequency <--> Place of origin	Local residents visit more frequently due to geographical accessibility.	[50]
Educational information <--> Place of origin	Chengdu has better educational resources, and the proportion of highly educated local residents may be higher.	[56]

), four hypotheses were proposed to guide the SEM construction:

Hypothesis 1 (Ha).

Sound intensity has a positive effect on subjective experience.

Hypothesis 2 (Hb).

Sound source recognition has a negative effect on subjective experience.

Hypothesis 3 (Hc).

Individual information has a significant effect on subjective experience.

Hypothesis 4 (Hab).

Sound intensity and sound source recognition significantly influence each other.

Based on these hypotheses, the SEM was constructed as shown in Figure 4.

The SEM was developed in AMOS Graphics 24, following the relationship assumptions mentioned above. SEM models typically use fit indices to evaluate the overall goodness-of-fit. Common indices include the Chi-square to degrees of freedom ratio (X²/DF), Goodness of Fit Index (GFI), Root Mean Square Error of Approximation (RMSEA), and Comparative Fit Index (CFI). Table 4 presents the fit indices for this SEM. The initial results indicated that X²/DF, GFI, RMSEA, and CFI did not meet the recommended thresholds, requiring modifications to the model. To optimize the model’s goodness of fit and ensure the scientific nature of the corrections, this study, based on acoustic theory, soundscape perception theory, and tourism behavior theory, selected and added 7 groups of residual correlations through the Modification Index (MI). The theoretical basis and literature support are summarized as follows (Table 5). After these modifications, all fit indices met the acceptable criteria (

Table 6. SEM Model Fit Indices (Before and After Modification).

Parameters	X2/DF	GFI	CFI	RMSEA
Criteria	<3	>0.9	>0.9	<0.08
Data values (before)	3.29	0.631	0.808	0.181
Data values (after)	1.62	0.9	0.921	0.075

).

3. Results

Through objective acoustic measurements, subjective questionnaire surveys, and multi-dimensional analysis, this chapter presents the spatial differences in soundscape sound source perception, sound intensity distribution, and subjective perception, and further reveals the interaction mechanism of soundscape perception in historical areas through SEM.

3.1. Analysis of the Spatial Distribution Characteristics of Sound Source Perception

This study utilized kernel density analysis and perceived occurrence (P_O) statistics to uncover the spatial distribution characteristics of natural and cultural soundscapes in the street and forest areas. Figure 5 shows the kernel density estimation based on 20 research points, displaying the potential high-value areas of sound intensity/sound source perception (the results are spatial interpolations of discrete points).

Overall, biological sounds had the highest perceived occurrence across the study area (P_O = 8.1), primarily consisting of bird songs, conversations, and other natural environmental sounds. Anthropogenic sounds ranked second (P_O = 5.7), including broadcasting and market-related sounds, while geophysical sounds had the lowest perceived occurrence (P_O = 3.2), mainly comprising wind and water sounds. Significant differences in sound source perception were observed: biological sounds were more frequently perceived within the street area than in the forest, whereas geophysical sounds were more prevalent in the forest than in the street.

Within the street area (Points 1–8), the dominant soundscapes were biological and anthropogenic sounds. Biological sounds had the highest perceived occurrence (P_O = 10.9), concentrated near the street entrance and main commercial areas. Anthropogenic sounds (P_O = 8.5) were predominantly observed in the vending zones and plazas, characterized by visitor conversations, vendor calls, and broadcasting sounds. Geophysical sounds were sporadically distributed in the street area, appearing near wooded or water features, but their perceived occurrence was relatively low (P_O = 2.3).

In the forest area (Points 9–20), geophysical and biological sounds prevailed. Biological sounds (P_O = 6.2) were primarily perceived in areas with high vegetation coverage. Geophysical sounds were concentrated around water bodies and adjacent areas, with a perceived occurrence of P_O = 3.81. Anthropogenic sounds (P_O = 3.79) were more scattered and less frequent than in the street, but still perceptible at high-traffic path intersections. In addition, the high concentration of geophysical sounds near water bodies, this spatial distribution may also be the reason for the poor correlation between geophysical sounds and other indicators.

At the interface between the street and forest areas (Points 1, 3, 4, 8), the perceived occurrence of biological and anthropogenic sounds were intermediate—slightly higher than in the forest but lower than in the street. Geophysical sounds were significantly less frequent compared to the forest area. The transition zone exhibited a notable complexity in soundscape composition and a relatively high soundscape diversity index (SDI), reflecting the coexistence of anthropophony (e.g., broadcast audio) and geophony (e.g., water sounds). However, this diversity was accompanied by reduced sound clarity, as overlapping and conflicting sources (e.g., loud announcements vs. soft wind sounds) diminished visitors’ ability to distinguish dominant sounds (Figure 5).

These findings indicate that cultural soundscapes (biological and anthropogenic) dominate the street area, while natural soundscapes (geophysical and biological) prevail in the forest. The interface between these zones is characterized by rich soundscape diversity but more complex and less distinguishable acoustic experiences due to mixed sound sources and higher visitor density.

3.2. Spatial Distribution Characteristics of Sound Intensity

Based on the kernel density analysis (Figure 5), overall sound intensity is higher within the forest than that within the historical district, while the boundary area generally has lower sound intensity compared to both the forest and the district. Specifically, the L_Aeq values are highest within the forest, especially concentrated around the water surface and near measurement point 8, followed by areas inside Jinli Ancient Street, such as the entrance. The L_Aeq values at the boundary between the district and the forest are relatively lower.

The L₁₀ values, which reflect peak sound pressure, are also higher in the forest than in Jinli Ancient Street, with lower average peak sound pressure shown at the boundary. Similarly, the L₅₀ values near the water surface and at point 8 in the forest are higher than at the entrance of Jinli Ancient Street, followed by the interior of the street, with the boundary having lower L₅₀ values. The same trend was observed in L₉₀ values, which are higher near the water surface and at point 8 in the forest moderate at the street entrance and interior, and lowest at the boundary. Finally, the L_10-90 value, typically indicating foreground sound or signal noise, is highest at point 8, followed by the performance area inside Jinli Ancient Street and the area near point 8 in the forest.

3.3. Spatial Variations in Subjective Perception

According to the 5-point Likert scale results, visitors rated the overall soundscape of the district as follows: pleasure at 3.6, comfort at 3.7, quietness at 3.0, and harmony at 3.6. The results of Spearman’s rho correlation analysis (Table 2) show a low correlation (r = 0.384) between harmony and pleasantness, indicating no significant association. Similarly, harmony showed weak correlations with both comfort and quietness, suggesting that its evaluation may be relatively independent and possibly influenced by individual interpretations of the soundscape. The strongest correlation was found between quietness and comfort (r = 0.909), which is consistent with the common understanding of the mass.

In terms of spatial distribution, the average sound evaluation inside the forest (3.67) is higher than that inside the street (3.21). The evaluations of pleasantness, comfort, quietness, and harmony of the forest are all higher than those of the street (

Table 7. Subjective evaluation differences between street and forest.

	Pleasant Level	Comfort Level	Quietness	Harmony	Total (Average)
Street	3.49	3.43	2.41	3.53	3.21
Forest	3.72	3.86	3.43	3.66	3.67

). Among them, the evaluation difference in quietness between the street and the forest is the largest. People generally think that the street is too noisy. However, despite the lower quietness score, the street still received relatively high ratings for comfort and pleasantness, likely due to the positive psychological effects of lively, socially engaging sounds such as human interaction and street activities. In addition, from the Spearman correlation analysis, it was found that the correlation between harmony and other indicators is poor. This may be due to individual differences in people’s understanding of harmony, as well as the fact that harmony shows relatively high scores in both forests and streets (which has certain differences from other subjective ratings).

The L_Aeq in the forest area is higher than that on the street, but tourists’ rating of the quietness of the forest (3.43) is significantly higher than that of the street (2.41). This contradiction may stem from the matching mechanism of “soundscape type—psychological expectation”. According to the theory of perceptual contrast in environmental psychology, the natural sounds in the forest (flowing water, bird songs) conform to tourists’ preset of tranquility and nature. Even if the sound intensity is high, they are still perceived as comfortable background sounds [51]. However, the artificial sounds on the street (vendors’ broadcasts, crowd noise) deviate from the expectation of “cultural immersion in historical blocks”. Even if the sound intensity is low, they are easily judged as “disturbing noise”. This is consistent with the conclusion of Guo et al. (2022) [55] that “tourists pay more attention to sound types rather than sound intensity in natural sound scenes”.

3.4. Structural Equation Modeling (SEM) Analysis

The modified SEM is shown in Figure 6, with the summary of regression coefficients and covariance coefficients listed in Table 6. As shown in

Table 8. SEM Path Regression and Covariance Coefficients.

Path			Estimate	S.E.	C.R.	p	Standardized Estimate
Sound perception	--->	Individual information	−0.013	0.015	−2.826	***	−0.195
Sound intensity	--->	Subjective perception	0.033	0.016	2.121	***	0.437
Sound perception	--->	Subjective perception	−0.076	0.036	−2.113	***	−0.574
Individual information	--->	Subjective perception	−0.727	0.378	−2.222	**	−0.36
Sound intensity	--->	LAeq	1				0.922
Sound intensity	--->	L10	1.382	0.167	8.274	***	1.017
Sound intensity	--->	L50	0.745	0.151	4.939	***	0.783
Sound intensity	--->	L90	0.57	0.15	3.8	***	0.695
Subjective perception	--->	Pleasant level	1				0.899
Subjective perception	--->	Comfort level	1.233	0.165	7.474	***	0.99
Subjective perception	--->	quietness	2.023	0.34	5.95	***	0.908
Sound perception	--->	biophony	1				1.047
Sound perception	--->	anthropophony	0.643	0.27	2.38	***	0.685
Individual information	--->	age	1				0.635
Individual information	--->	occupation	1.276	0.588	2.17	***	0.842
Individual information	--->	Visiting frequency	0.929	0.509	1.826	***	0.534
Individual information	--->	Educational information	−0.247	0.2	−1.235	***	−0.32
Individual information	--->	Place of origin	0.102	0.24	0.425	***	0.121
Individual information	--->	Monthly income	0.739	0.398	1.856	***	0.482
Sound perception	<-->	Sound intensity	2.147	6.935	1.165	0.269

** p < 0.01, *** p < 0.001.

, the hypothesized paths Ha, Hb, and Hc were all statistically significant. Specifically, Ha indicates that sound intensity positively influences subjective experience, suggesting that higher sound intensity levels are associated with more favorable acoustic perceptions. Hb reveals that sound source recognition negatively affects subjective experience, implying that increased awareness or identification of sound sources may reduce perceived soundscape quality. Moreover, Hb2 demonstrates that sound source recognition significantly influences subjective experience through the mediating role of individual information, highlighting the complex interaction between perceptual awareness and personal background characteristics. Hc shows that individual information has a significant negative effect on subjective experience, suggesting that certain demographic or behavioral traits are associated with lower soundscape satisfaction.

Further analysis of regression coefficients between individual information and its component variables indicates positive and significant relationships with age, occupation, visit frequency, and income, while negative and significant relationships were found with educational background and place of origin. This suggests that individuals who were older, had longer work experience, visited more frequently, and earned higher incomes tended to report lower levels of subjective experience. In contrast, those with higher levels of education and visitors from outside the area reported more positive evaluations of the soundscape. In addition, the results of the SEM model (Table 6) show that the standardized path coefficient of sound intensity on subjective perception is 0.437, and the factor loading of L10 (peak sound) (1.017) is significantly higher than that of L90 (background sound) (0.695), indicating that in historical areas, foreground sounds are more capable of dominating people’s perceptions.

According to Table 8, the relationship hypothesized in Hab did not show statistical significance, indicating that this path is not supported by the model. This phenomenon may be related to the division of geospace. Parks and neighborhoods have different dominant sound sources. High-intensity bird songs can enhance tourists’ satisfaction, while high-intensity artificial sounds and talking sounds only cause people to feel annoyed.

4. Discussion

4.1. Spatial Perception Differences and Coexistence Mechanisms of Natural–Cultural Soundscapes: The Role of Spatial Expectations

This study collected data on visitors’ sound source recognition, subjective experiences, and sound intensity using questionnaires and sound level meters. We propose that the influence of sound source type and sound intensity on subjective perception is significantly shaped by spatial context and cultural expectations—a dimension underemphasized in existing soundscape research on historic areas [49,55]. Tourists’ expectations regarding different spatial functions can amplify or buffer the impact of the acoustic environment on their subjective evaluations. For example, in industrial heritage sites, tourists’ expectations of mechanical sounds can enhance their sense of immersion [57]. This is consistent with the “Expectation–Perception Matching Theory” [49], but we have extended it to the context of cultural heritage. Based on the empirical evidence from the historical–natural interface zone, this study reveals two unique soundscape perception mechanisms, which supplement and refine the existing theories.

First, visitors’ context-specific expectations (quietness for forests, liveliness for historic streets) influence their subjective evaluations, potentially overriding objective sound intensity levels. For example, in our research and analysis, the L_Aeq were higher in the forest than in the street area, visitors rated the street as noisier than the forest. This contradiction can be explained by the expectation mechanism in environmental psychology, which refers to the positive impact of people’s expectations on their perceptions through “cognitive–emotional consistency” [58]. In the study area, tourists form a dual reference framework based on spatial functions: when entering a forest, they take “natural tranquility” as a reference (for example, expecting the sound of wind blowing through bamboo forests and distant water flows), so even high-intensity natural sounds are perceived as “consistent with the natural environment” and will not be labeled as “noise”. On the contrary, in historical streets, the reference shifts to “orderly cultural vitality” (for example, expecting folk performances and the cries of vendors), so suddenly appearing peak sounds (such as the sharp broadcast sounds of vendors) are regarded as “disrupting the expected order” and amplified as “noise”. This finding is similar to the results of the traffic noise reduction experiment conducted by Levenhagen et al. (2021) in U.S. national parks: reducing traffic noise levels did not significantly improve visitors’ soundscape pleasantness, and visitors’ evaluation of the soundscape depends on their pre-existing psychological expectations of natural sounds rather than the physical sound pressure level [59]. Liu et al. (2017) found in their study on the soundscape of natural parks that tourists’ tolerance for aircraft noise increased by 15% when they were informed in advance that “aircraft often pass by here [50]”. However, our study further specifies that this effect is strengthened in historic–natural interface areas—where natural sounds (e.g., wind through bamboo in Wuhou Shrine’s forest) are culturally framed as “complementary to historical ambiance”, whereas anthropogenic sounds (e.g., vendor calls in Jinli Street) are perceived as “disruptive to historical immersion” despite lower L_Aeq. In other words, while the street was objectively quieter, it was perceived as noisier—likely due to a mismatch between acoustic features and spatial expectations

Second, the difference in tourists’ perceptual evaluations between forests and urban blocks may be attributed to the composition of the dominant sound sources. According to the survey results, forest soundscapes are dominated by geophysical and biological sounds, while street soundscapes are dominated by biological and anthropogenic sounds. Based on this difference, this survey report shows that people can perceive a higher degree of pleasure, comfort, and harmony in forests. This indicates that in historical areas, people’s tolerance for high-intensity natural sounds may be higher than that for high-intensity cultural sounds. This finding is similar to the results of recent studies, mainly reflected in two aspects. On the one hand, researchers generally believe that natural soundscapes lead to higher satisfaction: a cross-cultural study by Purves and Wartmann (2023) shows that natural sounds are more often regarded as environmental backgrounds, while cultural sounds are to some extent considered invasive activities [53]; a study by Ye et al. (2024) also found that even if natural sound sources (such as the sound of flowing water) have a high sound pressure level (>60 dBA), they are still acceptable because they meet people’s expectations for the natural characteristics of historical environments [3]; at the same time, natural environments can reduce the annoyance caused by high sound pressure levels, thereby enhancing the pleasure of soundscapes [60,61]. Vegetation in urban green spaces can also soften sound perception, reduce negative emotions caused by loudness, and the reduction of human-related sounds in natural spaces can also improve people’s evaluation of soundscapes [3,62,63,64]. On the other hand, researchers believe that crowd noise is more likely to cause disgust: such sounds will destroy the spirit of the place [3]. In historical environments, human sounds, which are classified as biological sounds in this study but are essentially anthropogenic, are generally considered to have a negative impact on pleasure [44] because they destroy the authenticity of the historical soundscape that tourists pursue [49]; in addition, the density of street crowds may increase subjective loudness, which is consistent with the study by Meng et al. (2015) [65], and the dominant sound will affect tourists’ overall perception of the loudness of the space, and its impact is usually greater than that of background sounds or insignificant sounds [44]—for example, tourists in street areas regard conversation sounds and broadcast sounds as the dominant sound sources, while in forests, geophysical sounds (such as wind and water sounds) are often regarded as the dominant sound sources. Therefore, even if the sound pressure level in forests is higher than that in urban blocks, tourists still think that forests are quieter than urban blocks.

In conclusion, people’s expectations of soundscapes are shaped by their previous experiences in similar environments. When the actual soundscape aligns with these expectations, people are less likely to form negative impressions and may even be less aware of the existence of the sound environment [49,57]. For example, in studies, tourists rated the harmony, comfort, and pleasantness of street areas higher than their quietness. This indicates that despite the relatively high noise levels, tourists have certain expectations about the liveliness of the area, which helps to mitigate negative perceptions. Our research extends this insight to historical areas, showing that the “expected sense of liveliness” stems from cultural associations related to heritage (for instance, Jinli Street’s reputation as a “traditional commercial block”), making it a more important predictor of subjective experience than objective quietness.

4.2. Soundscape Conflict at the Historic–Forest Interface: Characteristics, Mechanisms, and Theoretical Implications

The Wuhou Shrine Museum’s forest (Cultural Relics and Western Areas) and historic district (Jinli Ancient Street) are directly adjacent, with multiple pathways on the north side facilitating sound interaction between the two spaces. Kernel density analysis of sound intensity, sound source recognition, and subjective experience revealed a critical phenomenon: the interface zone (Points 1, 3, 4, 8) exhibited higher soundscape diversity but lower subjective evaluations (comfort score = 3.1) compared to both the forest and the historic street. This high diversity-low satisfaction paradox defines the core of soundscape conflict at the historic–forest interface—a pattern not fully addressed in existing research on urban interface soundscapes [66].

In terms of sound intensity, the interface zone had lower L_Aeq than both the forest and the street, yet visitors reported more perceived noise (quietness score = 2.7, vs. forest 3.43 and street 2.41; Table 7). This discrepancy stems from sound masking by dominant anthropogenic sounds: peak sounds from the street (e.g., sudden vendor broadcasts, tourist chatter) masked subtle natural sounds in the forest (e.g., wind through bamboo, distant water flow), reducing visitors’ ability to perceive coherent soundscapes. This is consistent with recent studies, which show that anthropogenic noise can impair humans’ ability to recognize biological sounds and low-frequency sounds through spectral masking and cognitive interference [67,68]. This aligns with the study of Liu et al. (2014), who demonstrated that dominant sound sources in interface zones disrupt the detectability of complementary sounds, leading to perceived acoustic disorder [66]. Notably, the interface lacked shared activity nodes (e.g., plazas, green pavilions) to buffer sound transmission, unlike the forest’s water features or the street’s performance areas, which further exacerbated sound incoherence.

Regarding sound source recognition, the interface’s high SDI (driven by mixed geophony, biophony, and anthropophony) did not enhance perceptual richness but instead increased confusion. This finding is also consistent with that of Liu et al. (2014) [66], who found in their study that a higher SDI impairs the perception of bird songs (biological sounds). This is because anthropogenic sounds (e.g., street broadcasts, PO = 4.2) frequently overlapped with natural sounds (e.g., birdcalls, PO = 3.9) at the interface, creating “acoustic competition” that violated visitors’ context-specific expectations [66,67]. This finding extends the research of Schreckenberg et al. (2010) on landscape-soundscape interactions, which noted that “visual cues shape soundscape expectations, and mismatches intensify dissatisfaction” in heritage sites [69].

Subjective evaluations at the boundary are also lower than those on both sides. This phenomenon may be related to the influence of tourists’ visual perception on soundscape evaluation, and its underlying reasons can be further explained through the expectation mechanism [49,58]. Specifically, when the soundscape matches tourists’ cultural memories or experiential emotional memories, it triggers positive emotional responses; conversely, mismatched sounds break this connection, leading to negative evaluations. In the context of Wuhou Temple, tourists form two core emotional connections based on cultural cognition: (1) For the forest area (cultural relics area), the emotional connection is ‘historical tranquility’—tourists associate the forest with the cultural background of the Three Kingdoms, so natural sounds (such as the rustling of bamboo leaves in the wind) strengthen this connection and enhance comfort. (2) For the street area (Jinli Ancient Street), the emotional connection is ‘traditional liveliness’—tourists associate the street with ‘ancient commercial scenes’ (such as folk performances, traditional vendors’ cries), so moderate cultural sounds strengthen this connection and offset the negative impact of low quietness. However, in the interface area, mixed sounds (such as overlapping street broadcasts and bird songs) break these two emotional connections: artificial sounds disrupt the ‘historical tranquility’ of the forest, while faint natural sounds cannot support the ‘traditional liveliness’ of the street, ultimately resulting in a low harmony score. In addition, the high crowd density at path intersections also reduces tourists’ subjective evaluations [34]. It should be particularly noted that unlike the street area—where crowd noise is often regarded as “vibrant cultural vitality” [70]—the crowd sounds at the junction conflict with tourists’ expectation of a “peaceful transition between history and nature”, further exacerbating the evaluation contradiction.

Existing studies on historic or forest soundscapes often treat the two as isolated systems [31,45], but this study highlights that their interface is not a “neutral transition” but a dynamic zone of conflict shaped by three interrelated mechanisms: sound masking, audiovisual mismatch, and transient crowding. This found specifying that in historic–natural contexts, conflict arises not from high sound intensity alone, but from the “misalignment between sound composition, visual cues, and cultural expectations.” For example, the interface’s mixed sounds failed to align with either the street’s “cultural liveliness” or the forest’s “natural tranquility”, leaving visitors without a coherent perceptual frame. The revelation of the laws governing soundscape integration by this special type of space not only compensates for the limitation of existing studies that mostly focus on single spaces and ignore the complexity of boundary zones, but also provides a supplementary perspective of “cultural–natural composite scenes” for mainstream soundscape theories. For example, existing international studies are mostly based on the single logic of “natural soundscapes promoting restoration” or “cultural soundscapes enhancing identity”. However, this study, through case studies of boundary zones, confirms that the integration of these two types of soundscapes is not a simple choice between “natural sounds dominating” or “cultural sounds leading”, but rather requires building an appropriate integration model based on the “dual historical–ecological attributes” of the space itself. For heritage sites like Wuhou Temple, which are centered on the Three Kingdoms culture, the optimal state of soundscape integration in the boundary zone is “taking natural sounds (flowing water, bamboo forest sounds) as the base, and low-intensity cultural sounds (traditional musical instrument performances, soft explanations) as embellishments”. This not only retains the restorative value of the natural environment but also does not weaken the cultural immersion of historical scenes. This law not only provides a reusable theoretical framework for soundscape planning of similar “forest–historical boundary zones” (such as Suzhou gardens and surrounding urban green spaces, Kyoto’s historical districts and suburban mountains and forests) but also contributes empirical experience from the Chinese context to international soundscape research on heritage sites. It helps promote the improvement and deepening of the theory of “cultural–natural soundscape integration” in cross-cultural contexts, and further enhances the international academic attention and practical reference value of this study.

4.3. Factors Influencing Soundscape Perception: Sound Intensity, Sound Source Recognition, and Individual Characteristics

To explore how sound intensity, sound source recognition, and subjective experience interact, this study constructed a SEM to examine the effects of these variables on visitors’ perceptual outcomes. The results of the SEM (after modification: χ²/DF = 1.62, GFI = 0.9, CFI = 0.921, RMSEA = 0.075; Table 6) confirm that the three factors of sound intensity, sound source identification, and individual characteristics jointly explain subjective perception, and each has a different influence mechanism. Hypotheses Ha (sound intensity has a positive impact on subjective experience), Hb (sound source identification has a negative impact on subjective experience), and Hc (individual characteristics have a negative impact on subjective experience) all passed the significance test (Table 8). This result provides key empirical support for the multi-factor interaction research on soundscape perception and echoes and expands upon previous studies.

Hypothesis Ha (sound intensity has a positive effect on subjective experience) was supported. Among the sound intensity metrics (L₁₀, L_Aeq, L₅₀, L₉₀), the SEM regression coefficients showed a clear hierarchy: L₁₀ > L_Aeq > L₅₀ > L₉₀ (Table 6). This indicates that peak sounds have the strongest impact on subjective experience, while background sounds have minimal influence. This finding contrasts with Axelsson et al. (2010), who identified L_Aeq as the primary predictor of comfort in general urban spaces [71], yet aligns with the observation that “peak sounds dominate perceptual evaluations in heritage sites” because they disrupt the “historical immersion” visitors seek [3,44]. In addition, previous studies have shown that loud mechanical noises (frequently appearing on the street) can affect human auditory capacity, impede stress recovery, and lead to negative emotions [72,73,74]. This also explains why visitors perceive the street as noisier, as commercial streets often feature harsh, disruptive sounds such as vendor calls. For instance, even though the forest had higher L_Aeq than the street, the street’s more frequent peak sounds led visitors to rate it as “noisier”, highlighting that peak sound control, not just average intensity reduction, is critical for historic soundscape optimization.

Assume that Hab (there is a significant interaction between sound intensity and sound source identification) is not supported: the covariance coefficient between sound intensity and sound source identification is not significant (Table 8), which reflects the spatial segmentation of soundscapes in the study area. Historical streets are dominated by medium-to-low intensity anthropogenic sounds, while forests are characterized by high-intensity natural sounds. As found by Liu et al. (2014), when soundscapes are spatially separated (for example, separated by the bamboo forest buffer zone of Wuhou Temple), intensity and sound source type become “decoupled”—this explains why no direct interaction was observed [66]. This finding adds nuance to the model by Aletta et al. (2016), which assumes a universal interaction between intensity and sound source identification [17], whereas this study shows that this relationship depends on the spatial context. Therefore, the value of the non-significant result of Hab lies in that it reveals the spatial boundary conditions of the “sound intensity–sound source identification association”—when there is an obvious spatial separation of soundscapes, the interaction between the two becomes ineffective, which has certain significance for revising the universal model of soundscape perception.

In addition, Hypothesis Hb, which examines the effect of sound source recognition (including anthropogenic and biological sounds) on subjective experience, shows that sound source recognition (P_O) has a negative impact on subjective experience, with biological sounds having the most significant effect. This result differs from our initial assumption that biological sounds, particularly non-human biophony like bird songs, would exert a neutral or positive influence on subjective experience, as supported by prior studies linking natural biophony to restorative effects [32,33,34,35,36]. The key explanation lies in the composition of “biological sounds” in our dataset: while bird songs were present, they were outnumbered by human conversations (classified as biological sounds in our framework but inherently anthropogenic in nature). Biological sounds, such as conversations, have a notable negative impact on visitors’ subjective experience when they frequent occurrence [3,70]. Studies have found that human-dominated biological sounds have destroyed “historical authenticity” [70]. The kernel density analysis (Figure 5) also reveals that areas with higher perception of biological sounds tend to have lower subjective ratings. In the forest, although there are positive sound sources such as the chirping of birds, the human conversation has a strong masking effect, which may be the reason why the high occurrence of biological sounds leads to a decrease in people’s subjective experience. Therefore, in forests located in historical areas, it is necessary to control the flow density of people to a certain extent and set up quiet visiting areas [38], so that biological sounds such as bird calls can be more distinct.

Hypothesis Hc (individual information has a significant effect on subjective experience) was supported: Individual characteristics exerted a significant negative overall effect on subjective perception (Table 8), but subgroup differences revealed critical nuances tied to the study’s questionnaire data (Section 2.4). Age correlated positively with negative perceptions: older visitors were more sensitive to peak sounds, as Schreckenberg et al. (2010) noted, “noise sensitivity increases with age due to auditory system changes” [69]. Visit frequency also mattered: frequent visitors reported lower comfort, likely because they had higher expectations for “acoustic consistency”. In contrast, education and place of origin had positive associations: visitors with graduate degrees or above showed greater tolerance for “controlled cultural noise”, while non-local tourists perceived street liveliness as a “cultural attraction” rather than a disturbance. Income also had a weak positive effect, with higher-income visitors more likely to overlook minor noise disruptions.

Furthermore, this study’s unique regional, site, and cultural backgrounds enable its findings to supplement international soundscape theory and offer differentiated contributions in three aspects: (1) Regional dimension: Unlike studies on low-density European heritage sites [44,45,46], Wuhou Temple (in the core of Chengdu, a western Chinese megacity) features “high overlap between historical space and urban life”. High visitor density in Chinese scenic spots creates frequent peak sounds (e.g., folk performances, tourist conversations), making L₁₀ (not the commonly used L_Aeq) the core indicator of subjective experience, which informs research on high-density Asian urban heritage sites. (2) Site dimension: Different from Aletta et al.’s (2016) open European urban parks [70], Wuhou Temple adopts a “closed-separated” layout via “bamboo buffers + low walls”. Soundscape overlap concentrates at the junction of historical streets and urban forests, leading to an insignificant “sound intensity–sound source identification” interaction—contradicting the “universal positive interaction” assumption in international models and providing a classification reference for site-specific soundscape studies. (3) Cultural dimension: Contrasted with natural sound protection in nature reserves [72], Wuhou Temple (a core Three Kingdoms cultural heritage site) endows forest sounds (e.g., bamboo wind, water) with cultural meanings, giving them dual “ecological restoration-cultural carrying” functions. This expands soundscape theory and offers a non-Western perspective for natural soundscape protection in cultural heritage sites. In summary, this study verifies the applicability of international soundscape theory in specific scenarios, reveals soundscape laws of high-density, culturally overlapping East Asian heritage sites, and provides key empirical support for the regional and cultural diversity of international soundscape research.

To contextualize the study’s methodological and theoretical contributions, this research develops a tripartite soundscape framework tailored explicitly for historic districts—a tool designed to address the unique interplay of culture, nature, and human perception in heritage settings. This framework consists of three core components: (1) sound source typology, (2) perceptual dimensions (pleasantness, quietness, harmony, comfort), and (3) socio-demographic profiling. This framework differs from standardized soundscape scales by prioritizing context-specificity over universality. For instance, Axelsson’s model focuses on a universal Pleasure-Eventfulness-Familiarity triad [71], and ISO/TS 12913-2 emphasizes general spatial metrics [75], yet both of these tools overlook the “cultural compatibility” of sounds in heritage sites. In contrast, our framework explicitly includes “harmony” as a key perceptual dimension, defined as the alignment between acoustic elements and a site’s historical identity. In practice, “harmony” is operationally defined through survey questions such as “Does the sound of this area match its historical characteristics?” and measured using a 5-point Likert scale. This dimension was identified via exploratory factor analysis (Section 2.4), which showed that harmony loaded separately from pleasure or comfort, capturing unique variance related to perceived congruence between acoustic elements and the site’s cultural identity. This allows us to distinctively quantify how sound combinations align with heritage expectations. In addition, compared to objective bioacoustic indices (e.g., NDSI for biodiversity quantification) [76], our subjective classification captures visitor-perceived sound dominance. Socio-economic variables (income, origin) extend beyond the SSID protocol’s basic demographics [77], revealing subgroup variations (e.g., income-correlated quietness preferences). Future studies could integrate ISO-compliant acoustic indices (e.g., NDSI/ACI) with our perceptual framework to enhance ecological validity while retaining contextual strengths, or adopt machine learning for cross-cultural predictive modeling.

5. Conclusions

This study examined Jinli Ancient Street in Chengdu, China, using a combination of field measurements and visitor surveys to investigate the interactions between natural and cultural soundscapes in a historical urban setting. The research employed kernel density analysis, correlation analysis, and SEM to explore how sound intensity and sound source recognition influence subjective soundscape experiences. The key findings are as follows:

(1) Impact of Sound Intensity: High sound intensity (L₁₀) had the greatest influence on subjective experience, while low sound intensity (L₉₀) had a weaker effect. This highlights the importance of managing high-intensity or peak noise sources to prevent disruptions to visitors’ overall soundscape perception.

(2) Conflict Between Natural and Cultural Soundscapes: The soundscape diversity at the boundary zone—combined with mismatched audiovisual cues and elevated crowd density—contributed to a more fragmented and less satisfying acoustic experience. Future interventions should aim to alleviate these conflicts through integrated design strategies and visitor flow management, facilitating more harmonious coexistence of natural and cultural sound elements.

5.1. Theoretical Implications

To address gaps in historic soundscape theory [13,57,63], this study makes three novel, previously absent contributions:

(1) Extending the expectation–perception framework to heritage contexts: It identifies an “expectation–audiovisual matching” mechanism, showing that soundscape perception in historical interfaces depends not only on functional expectations (e.g., “forests = quiet”) but also on cross-sensory consistency with cultural identities (e.g., Three Kingdoms heritage in Wuhou Shrine).

(2) Developing a culture-adapted perception dimension: Exploratory Factor Analysis (EFA) verifies “harmony” (alignment between sounds and historical identity) as an independent perceptual dimension (explaining 15.951% of variance), filling the gap in standard tools (e.g., Axelsson’s triad, ISO/TS 12913-2) that lack cultural compatibility assessment.

(3) Revealing spatial boundary conditions for soundscape interactions: It confirms that the “sound intensity–sound source correlation” assumed in universal models disappears when soundscapes are spatially segmented (e.g., street vs. forest), highlighting “spatial segmentation” as a critical moderator for interface soundscape research.

In sum, this study takes the interface of “historical streets–urban forests” in Wuhou Temple as a case, and for the first time quantitatively reveals the contradictory mechanism of “high soundscape diversity but low subjective satisfaction” in such spaces. It confirms that the triple interaction of “sound masking + audio-visual mismatch + transient crowding” is the core inducement of soundscape conflicts in the interface area, providing a reusable empirical analysis framework for soundscape research in similar historical–natural interleaved zones.

5.2. Limitations

This study has several limitations that should be acknowledged:

Site and temporal constraints: Findings are specific to Wuhou Shrine (unique Three Kingdoms context, compact layout) and limited to weekday mornings (April 2021), excluding seasonal/diurnal soundscape variations.

Sample and classification gaps: The 113 valid responses underrepresent groups like international tourists; human biophony (e.g., conversations) was not distinguished from non-human biophony (e.g., birdcalls), and expert-naive participant perception differences were unanalyzed.

5.3. Policy Recommendations

These research findings provide actionable insights for soundscape management and sustainable development in historical–forest transition zones. They are applicable to urban environments where historical areas coexist with natural spaces, such as urban heritage parks and urban cultural parks. Based on the findings of this study, we propose specific and implementable strategies from three levels:

First, adopt a “space-time-sound source” 3D zoning system. Spatially, divide areas into “cultural sound-dominated zones” (e.g., Jinli Ancient Street core, protect folk performances, ban electronic loudspeakers), “natural sound-dominated zones” (e.g., Wuhou Shrine relics forest, limit gatherings to ≤20 people, conversation volume <50 dBA), and “transitional zones” (e.g., interface points 1, 3, 4, 8, set ≥15 m bamboo-water buffers to attenuate sound). Temporally, schedule cultural activities (e.g., folk performances) in cultural zones 9:00–11:00 to avoid forest morning exercise peaks (7:00–8:00). Classify sound sources by prioritizing bird songs/water flows, restricting human conversations, encouraging traditional instruments, and prohibiting modern construction—integrated into local Heritage Site Soundscape Protection Regulations.

Second, optimize quiet zones with “natural sound enhancement + artificial sound shielding.” Amplify natural sounds via water/wind guiding devices near forest water bodies and mixed vegetation (ginkgoes + bamboo) to extend bird songs. Block artificial sounds with composite barriers (1.2 m soil slopes + 2 m evergreen hedges) at quiet zone boundaries.

Third, apply digital simulation for “pre-evaluation—dynamic monitoring.” Pre-plan: build a soundscape model, input pedestrian density/vegetation coverage, and simulate L10/LAeq distribution to identify conflict hotspots. Monitor in real time: deploy sensors at 20 points, and warn when L10 > 65 dBA; combine with a tourist APP’s 5-point rating for “objective + subjective” dual monitoring.

These strategies enhance visitor experiences, preserve dual values, and promote sustainable coexistence of historical and natural soundscapes.