Sound masking of residential noise by a birdsong depending on spectral and temporal characteristics Songmi Lee 1 , Chunwon Eom 2 , Jongkwan Ryu 3 Chonnam National University 77, Yongbong-ro, Buk-gu, Gwangju, Republic of Korea

ABSTRACT This study aimed to investigate the annoyance reduction of residential noise by adding a birdsong. First, seven noise sources were selected for the experimental stimuli as residential noise according to spectral and temporal characteristics. Two types of birdsong sound sources with different spectral characteristics were selected among collected birdsongs. The experiment evaluates the annoyance of single residential noise, birdsong, and combined (residential noise+birdsong) sound source using a 7 verbal-points scale. Also, this study analyzed physical and psychoacoustic parameters. As a result, the annoyance reduction by birdsong, which dominates the high-frequency sound, was great for residential noises with large L Ceq -L Aeq value. This is due to the effect of informational masking because the residential noises (maskee) have different frequency characteristics with birdsong (masker). In addition, the relationship between annoyance reduction and temporal characteristics and psychoacoustics parameter was analyzed. 1. INTRODUCTION

Residents' complaints about various residential noise, including floor impact noise in apartment buildings, have been steadily increasing. In order to reduce complaints, many studies have been conducted on technical efforts such as floor insulation performance. However, there are still many complaints, it needs to solve this problem, other methods such as masking technology were needed. The previous researches have investigated influence of adding natural sound (birdsong, water sound etc.) on soundscape with outdoor road traffic noise [1~4] and construction noise [1, 4] trough laboratory experiment. However, there was a difference between indoor and outdoor environments, and various noises occur not only from the outside, but also from neighboring units in the muti-story house like apartment. It was reported that subjective response of floor impact

1 smi4879@naver.com 2 ucw1472@naver.com 3 jkryu@jun.ac.kr

noise such as child’s running (low frequency dominated) and chair scrapes (high frequency dominated) was not affected by indoor water sounds on overall comfort [5]. However, there is a lack of research about the effect of adding natural sound on subjective response of various residential noise source. This study conducted to understand annoyance reduction of various residential noise by adding birdsong on the basis of spectral and temporal analysis. 2. METHOD

2.1. Stimuli For the experimental stimulus, actual residential noises were used, and in case of outdoor noise, they were filtered out by the average value of façade insulation of a residential building [6]. This study analyzed spectral and temporal characteristics used AS-70 software (RION). To select sound source for experimental stimuli, k-means clustering (elbow method) was first performed using Python with spectral (1/1 octave band, 31.5~2k Hz) and temporal (0.1~9.993 s, 125 ms) data of residential noise. Figure 1 indicates three the spectral cluster (k=3) as function of 1/3 octave band. A total of 7 noise types were then selected, and the selected stimuli were divided into three groups of spectral ( L Ceq - L Aeq ) and temporal ( L A10 - L A90 ) characteristics such as impact noise, fluctuating noise, stationary noise. Meanwhile, birdsong sounds were collected from National Institute of Biological Resources [7], and two birdsongs (Cuculus canorus, Phoenicurus auroreus) with different spectral characteristics were selected among various birdsongs.

iebjonpinins!

Figure 1: The result of k-means clustering of residential noise based on spectral characteristic

2.2. Experiment Design For the experiment, residential noises such as floor impact noise (child jumping, child running) conversation, air-conditioner outdoor unit, piledriver, plumbing and road traffic noise were used. The duration of the noise was 10 seconds and sound pressure level referred to previous research [8] and similar to Korea indoor noise criteria. Floor impact noise and piledriver were L Amax of 55 dBA and L Aeq of 50 dBA for others. Also signal-to-noise ratio (SNR) of residential noise (noise) and birdsong (signal) were 0 dB. As shown in Figure 2, both sounds of Figure 2(A) and Figure 2(B) were dominated by low frequency, but floor impact noises have higher sound pressure level than the conversation and outdoor unit. And Figure 2(C) plots that sound pressure level in low frequency was similar to middle and high frequency. On the other hand, in case of birdsongs, both canorus (650 Hz) and phoenicurus auroreus (5k Hz) were dominated at high frequency in Figure 2(D). The experiment was conducted in anechoic chamber (H x W: 2.7 m x 4.4 m = 11.9 m 2 ) and stimuli were presented in random order using woofer speaker (Dynaudio BM14S) and headphone (Sennheiser HD 600). The experiment consisted of two sessions: Experiment 1 was for single residential noises and experiment 2 was for combined sounds (residential noise + birdsong). For the evaluation method, a Korean 7 verbal-point scale [9] for annoyance as “ Assume that you are resting in the living room and listen to the following sound and evaluate your annoyance due to the sound”. A total 31 subjects (24 men and 7 women) with self-reported no hearing impairment were volunteered, and the average age was 23.3 years (SD: 2.3 years).

‘Sound pressure level [dB] 100 0 oo 50 0 30 » 10 area BLS 63 125 250 500 Tk 2k Aff octave band frequency{lia)

Figure 2: Spectral characteristic of residential noise (A, B, C) and

birdsong (D) of experimental stimuli as a function of 1/3 octave

band, 3. RESULT

3.1 Annoyance of single residential noise Figure 3 illustrates the annoyance of single residential noises. Impact noise such as child jumping (5.9), child running (6.3), piledriver (5.5) showed higher annoyance and road traffic (3.2) showed the lowest annoyance. As shown in Table 1, the relationship between the annoyance of single residential noise and acoustic parameters were analyzed. For impact noise (floor impact noise and piledriver), which showed the highest annoyance, has low the equivalent sound level ( L Aeq ), but high L A10 - L A90 . Also, the L Ceq - L Aeq of floor impact noises were higher than the piledriver. On the other hands, conversation and outdoor unit with large L Ceq - L Aeq have also high annoyance at the same L Aeq level. The relationship between single residential noise and annoyance was higher correlation coefficient in L Ceq - L Aeq (r=0.72) than L A10 - L A90 (r=0.68). It was found that the subjects

evaluated sound with high sound pressure level in the low frequency band and high fluctuation as more annoying sound.

Figure 3: Annoyance of single residential noise

Table 1: the relationship between the annoyance of single residential noise and physical acoustic

parameters

Stimuli L Aeq [dB]

L Amax

L Ceq – L Aeq

L A10 - L A90

[dB] Annoyance

[dB]

[dB]

Child jumping

47.7 55.1 31.8 27.0 5.9

Impact

Child running 49.7 55.1 30.2 14.9 6.3

noise

Piledriver 47.4 55.1 6.9 15.8 5.5

Conversation Fluctuating

50.1 56.1 11 12.1 5.2

noise

Plumbing 50.0 55.4 5.5 12.1 4.2

Outdoor unit Stationary

50.0 52.2 14.6 2.7 4.9

noise

Road traffic 50.1 52.9 7.8 3.2 3.2 Correlation coefficient with

annoyance -0.50 0.48 0.72 0.68 1

3.2. Annoyance of combined sound (residential noise and birdsong) Figure 4 indicates the annoyance of combined sound with residential noise by adding birdsong. The annoyance was reduced in most residential noises, except of some sounds. In particular, a bird song sound (phoenicurus auroreus) showed higher reduction. However, in the case of piledriver and road traffic noise, the reduction of annoyance was rather increased or insignificant due to the two birdsongs, and the plumbing was reduced only by the cuculus canorus. Paired T-tests were conducted through SPSS (IBM, ver 27) for the reduction of annoyance by adding of birdsong to a single residential noise source. As a result, it was found that two birdsongs were significant in the floor impact sound and outdoor unit noise. In addition, although two birdsongs were reduced in conversation, it was found to be significant only for phoenicurus auroreus, and the annoyance reduction of outdoor unit was significant in two birdsongs. This result suggests that the subjects were rated with different evaluation ratings between single residential noise and combined sounds.

Annoyance [score], Child ——Child_—-Conversation Outdoor unit Piledriver Plumbing Road traffic jumping running. nr

Figure 4: Annoyance of single residential noise and combined sound (CC: Cuculus

canorus, PA: Phoenicurus auroreus, **: p <0.01, *: p <0.05)

4. CONCLUSIONS

This study investigated the annoyance reduction depending on a spectral and temporal characteristic by adding birdsong into residential noise sources. As a result of the experiment, even though the equivalent noise levels ( L Aeq ) of floor impact sounds were low, their annoyance was great because of high L Ceq - L Aeq and L A10 - L A90 value. It was also observed that the annoyance of combined sound (residential noise + birdsong) was decreased in most sound sources regardless of temporal variation characteristics. In particular, it was found that the annoyance by adding two birdsongs were reduced for the sound source with the large sound energy in the low frequency band. In addition, it was found that the annoyance by adding a birdsong which has a peak of sound in higher frequency between two birdsongs, was relatively significantly reduced. According to the present analysis, spectral and temporal characteristics had little effect on annoyance reduction of residential noise, which was different from the ratings for single residential noise evaluation. It was considered that the informational masking by the asymmetric spectrum may have had an effect. In the future, it is necessary to investigate that annoyance reduction by adding birdsong using additional parameters (e.g., spectral centroid) including psycho-acoustical parameters. 5. ACKNOWLEDGEMENTS

This work is supported by the Korean Agency for Infrastructure Technology Advancement (KAIA) grant funded by the Ministry of Land, Infrastructure and Transport (Grant 21CTAP- C163631-01). 6. REFERENCES

1. Jeon, J. Y., Lee, P. J., You, J., & Kang, J. Perceptual assessment of quality of urban

soundscapes with combined noise sources and water sounds. The Journal of the Acoustical Society of America, 127(3), 1357-1366. (2010).

Annoyance [score] Child jumping Child running Single noise © Combined_CC &Combined_PA Ih Piledriver Conversation Plumbing Outdoor unit Road traffic Maskee type

2. Coensel, B. D., Vanwetswinkel, S., & Botteldooren, D. Effects of natural sounds on the

perception of road traffic noise. The Journal of the Acoustical Society of America, 129(4), EL148-EL153. (2011). 3. Hao, Y., Kang, J., & Wörtche, H. Assessment of the masking effects of birdsong on the road

traffic noise environment. The journal of the Acoustical Society of America, 140(2), 978-987. (2016). 4. Hong, J. Y., Ong, Z. T., Lam, B., Ooi, K., Gan, W. S., Kang, J., ... & Tan, S. T. Effects of

adding natural sounds to urban noises on the perceived loudness of noise and soundscape quality. Science of the Total Environment, 711, 134571. (2020). 5. Yang, W. Effects of Indoor Water Sounds on Floor Impact Noise Perception and Overall

Environmental Comfort. Journal of The Korean Society of Living Environmental System, 25(5), 611-619. (2018). 6. Ryu, J., & Song, H. Effect of building façade on indoor transportation noise annoyance in

terms of frequency spectrum and expectation for sound insulation. Applied Acoustics, 152, 21- 30. (2019). 7. National Institute of Biological Resources, https://species.nibr.go.kr/home/mainHome.do?cont_link=012&subMenu=012005&contCd=0 12005002 8. Park, S. H., Lee, P. J., & Lee, B. K. Levels and sources of neighbour noise in heavyweight

residential buildings in Korea. Applied Acoustics, 120, 148-157. (2017). 9. Ryu, J, Jeon, J, Kim, H. Development of Noise Annoyance Scale and Criteria of Residential

Noises through Auditory Experiments.Transactions of the Korean Society for Noise and Vibration Engineering,15(8),904-910. (2005).