4.1. Soundscape Features of the Urban Recreational Forest Parks
People’s perceptions of dominant sounds and non-dominant sounds in city parks were easily detected and showed a large uniformity among participants, confirming the study by Liu et al. (2018) [
39]. Dominant sounds in city parks included music, surrounding speech, children playing, and cars. In contrast, submissive sounds included a train and dogs. Although some individual sounds such as instruments, lawn mowing, ducks, diabolo, engines, and motorbikes were not frequently perceived, they were perceived as loud sounds when they were detected. The most preferred sound was that of a bird’s song, but it did not belong to the group of dominant sounds. Instead, the dominating sounds of square dancing, surrounding speech, and automobiles in the city parks were easily detected and were perceived as being very annoying by the majority of the respondents. The finding that participants exhibited higher preferences for natural sounds and lesser preferences for artificial sounds is in line with previous research [
17,
18,
46].
If the purpose is to increase the positive perception within parks for the majority of visitors, these findings should be taken into consideration. One suggested way to increase the positive experiences of park attendees is to increase the occurrence of bird songs either by adding a variety of habitats for birds or by adding artificial bird songs to the environment [
10,
46,
57]. However, simply improving habitats or adding bird songs does not reduce other sounds negatively perceived as noise. Studies showed that masking noise by adding bird songs might improve the soundscape perception [
11]. Still, however, the negatively perceived noise cannot be too loud if bird songs are to be masking it [
58]. Furthermore, there is an interlinkage between vision and sound wherein a dissatisfaction of visual features such as cars affects soundscape perceptions negatively [
22,
43,
59]. One potential advancement could be to improve visual features or build visual barriers towards these sounds.
Perceived loudness (OPL), annoyance (OPA), and preference for sound (OPR) were not perfectly matched with each other. Nearly 86% of respondents were not averse to the soundscape (from the point of overall soundscape annoyance). Yet, only 58% of them were satisfied with the soundscape (OPR). Thus, not being annoyed or averse to a soundscape does not equate to satisfaction. More attention should thus be paid to enhancing sounds that people prefer, while at the same time exploring different people’s specific preferences for the soundscape in urban recreational forest parks.
4.2. Soundscape Perceptions and Preferences among Different Social, Demographic and Behavioral Park Users
This study recognized that almost every sound, as well as the overall soundscape, were perceived differently in five different dimensions.
Age and Familiarity of site was the most influential dimension on perception of and preference for individual sounds (with 28 perception parameters of certain sounds). Elderly individuals who were familiar with the environment tended to perceive most natural, musical, and artificial sounds infrequently, while at the same time, more fully enjoyed these sounds when they were heard. This is consistent with the result that long-term experiences in particular locations could reduce the sensitivity of the acoustic environment [
15]. As the elderly tend to visually perceive parks as a natural habitat and enjoy the parks more than the younger population [
60,
61], this finding could also be related to the notion that the elderly have stronger connections to the area, which is linked to the theory of place identity or place attachment, which are in turn linked to greenery [
62]. For example, in the study conducted by Hedblom et al. (2017), older people reported having a stronger experience of the site related to the sounds of nature and felt calmer as a result of the sound of rustling trees and bird songs than middle-aged and younger people [
63].
However, it is interesting to find that elderly individuals who were familiar with the environment tend to perceive musical sounds more frequently. It may be because square dancing and singing were the most dominant sounds, so the people who were familiar with the soundscape would tend to perceive them easily. Another reason might be the elderly who were familiar with the environment tend to be the ones participating in square dancing and singing. Notably, Age and Familiarity of site was the only dimension of the attributes that influenced the perceived visual sources (VS). Thus, the elderly who were familiar with the environment more easily discovered the visual sources of natural and musical sounds [
64]. It is worth mentioning that although a variety of individual sound perceptions were affected by Age and Familiarity of site, overall soundscape perceptions were not influenced at all. We interpret these data as supportive of the idea that the Chinese elderly who were familiar with the environment had a high preference and low sensitivity towards individual sounds, while square dancing and singing were two special sounds which largely aroused their attention.
Respondents with a higher level of education and a greater level of income perceived the overall soundscape to be more annoying and showed less satisfaction than those who were less educated and had lower incomes. Specifically, they had lower preferences for people singing. This finding is consistent with previous studies that showed that the higher the social status is, the less tolerance there is for the soundscape [
19]. Notably, age and familiarity affected preferences for natural sounds while education and economic status did not, except for low preferences for dogs. This is perhaps because individuals with higher education and higher income also had high expectations for natural sounds, while the sounds of dogs, to some extent, were out of their realm of expectation [
48]. This might also be linked to the finding that they perceived a train and cars more frequently. Overall, we interpret these data as supportive of the notion that although people with a different Educational and Economic condition had varying opinions towards sound sensitivity, it is clear that people with a high Educational and Economic condition generally showed a lower tolerance for soundscape.
As for the Companion and Type of recreational use dimension, previous research showed that companions seldom influenced soundscape perceptions and preferences [
47,
54]. However, in this study, it was surprising to find that Companion and Type of recreational use had influences not only on the perceived occurrences (PO) but also the loudness (PL) of artificial sounds, particularly surrounding speech and playing children. Namely, the respondents with more companions also preferred certain activities (such as parent-child activities, fitness and health activities, sports and leisure activities, or social activities) and perceived sounds both loudly and frequently. One possible reason could be that sounds might originate from participants’ companions or their engagement in activities, making them easier to experience or recognize. In addition, respondents with more companions perceived the train and automobile sounds more frequently. This may be because these sounds would interrupt their conversations or any interactions with other people [
19]. Perhaps it is also why these participants also perceived overall soundscapes to be much more annoying and louder. Interestingly, the perceived occurrences (PO) of diabolo showed an adverse trend compared to other artificial sounds. Perhaps this is because diabolo, as a specialized activity, was usually played by no more than two people (regarded as fewer companions), and people who participated in this activity could detect the sounds of diabolo more easily than others. Overall, we interpret these data as supportive of the idea that the Companion and Type of recreational use factor is highly correlated with the perception of artificial sounds, while its influence was rarely reflected on sound preference.
As for the Gender dimension, this study found that the perceived loudness of overall soundscape (OPL), perceived occurrences (PO), and preference (PR) of several sounds can be influenced by the Gender factor. This finding is in line with Hedblom et al. (2017), whose study found differences between gender wherein sounds of nature were linked to bird species experiences [
63], but not in line with previous studies which argue that Gender had no influence on soundscape perception [
15]. Furthermore, in this study, females tended to perceive the overall soundscape as being louder, while at the same time, were more sensitive to the sounds of insects and showed less tolerance of the sounds of playing children and that of engines. It is perhaps because females are more sensitive to sound than males [
65], and easily recognized certain sounds that men did not discern. Another reason is that women, who usually take children to urban recreational forest parks, were closer to this sound source, causing them to experience a greater degree of annoyance [
19]. Overall, we interpret these data as supportive of the idea that females tend to show high sensitivity and low tolerance towards certain sounds.
Finally, this study showed that Length of stay was more correlated with the overall soundscape preference (OPR) and the perceived occurrences (PO) of individual sounds, which is consistent with previous studies [
15,
18]. The longer the respondents spent in the parks, the more satisfaction they had about the overall soundscape [
66]. Alternatively, it can also be considered as the better the soundscape is, the more (time) respondents wish to stay in the park. The results indicate that longer stays correlate with greater sensitivity towards birds and broadcast music and less sensitivity towards cars and motorbikes. Overall, we interpret these data as supportive of the idea that Length of stay is a special attribute, which not only positively contributed to overall soundscape preference but also had an adverse impact on perceptions between the favored sounds and disliked sounds.
Above all, the five dimensions of factors showed different degrees of importance on various perspectives indicating the importance of the impact of personal attributes on soundscape perception in urban recreational forest parks. Designing a satisfying soundscape thus needs to meet various expectations of people in the five dimensions.
The urbanization is continuously rapid in China and to highlight some findings might increase the overall well-being for urban citizens with different backgrounds. In terms of individual sound perceptions and preferences, this study found that Age and Familiarity of site showed the greatest influence compared to other dimensions. We have not measured the frequency of different age groups in the park, but nevertheless, urban forest recreational parks in China are important for the elderly population. Due to rapid urban development, many forest parks are rather new in China, and so are those that this study surveyed in Xi’an (all except for one were less than 12 years of age). Thus, as people are becoming more settled and become more familiar with these parks, one could suppose that familiarity, such as sense of place and place attachment, will increase in the future, further showcasing the importance of highlighting the findings in this study. It is also worth noting is that older forest parks in China traditionally have walls around them, reducing the noise from nearby car traffic, while newer parks, in general, do not. Thus, the newly established parks do not have walls. Perhaps there will be a need to return to the older, traditional ways of building parks with walls due to the need for artificial noise barriers, which could be designed to be green walls [
67]. Furthermore, Age and Familiarity of site was the only dimension highlighting a positive preference for bird and insect soundscape. One advancement would be to provide more space and habitats to allow for a greater number of bird and insect species. Increasing positive natural sounds might not only increase positive sound perceptions but could also be linked to increased health or stress reduction [
58].
With a rapidly increasing middle class in China and increasing levels of higher education [
68], future demands for parks might change. Our findings showed that people with higher education levels and greater income had a lower tolerance towards traffic noise and songs. Thus, to satisfy a growing urban middle class in need of restoration who are to a higher degree more annoyed with certain sounds than other demographics, there is a need for a new take in responsive and thoughtful park planning of soundscape.
The results further showed gender differences. It is a relatively new field of knowledge of how different sexes use and perceive urban forest parks [
29]. It might be due to men and women using different parts of the forest parks, and thus they are exposed to different sound sources [
69]. This might be taken into consideration when planning for soundscapes.
Although all respondents had high tolerances towards sounds, most were not satisfied with what they heard. The results indicated that depending on age, familiarity, gender, and other factors, people had different expectations and preferences for sound. One possible solution could be to dedicate specific areas for relative activities for certain user groups to attenuate aural dissatisfaction in highly visited Chinese forest parks.