Soundscape design for a residential development: a case study Seena Sajeev 1 Rebecca Romeo Pitone 2 Danuta Nabiałek 3 Julien Batten 4

Jack Harvie-Clark 5

Apex Acoustics Ltd Design Works, William Street, Gateshead NE10 0JP

ABSTRACT A soundscape approach was used for a co-housing development in Newcastle upon Tyne, UK. The outcomes are compared and contrasted with a conventional noise impact assessment. In-situ soundwalks were conducted with the co-housing group during daytime and night-time periods, covering locations with varied sounds and future uses. A virtual soundwalk, including external and internal scenes (with windows open and closed), was created using binaural recordings from the in- situ soundwalk. A separate virtual soundwalk event was organised for the public to investigate the ecological validity of the auralisations. Data collection and analysis methods followed ISO 12913. The questionnaire responses revealed quite different assessments of the acoustic environment compared with the measured decibel values of L Aeq, T . These differences could be explained by the perception of the acoustic environment. The process helped the co-housing group to identify their expectations and preferences of the sound environment of their future homes. Ideas to encourage and highlight positive (natural) sounds were identified, as were different uses for external areas. The study illustrates the potential and challenges of using a soundscape approach for residential developments, highlighting the aspects of co-creation and inclusivity. 1. INTRODUCTION

A co-housing development, named Co-HUT (Co-Housing upon Tyne), has been proposed in Newcastle upon Tyne. It comprises 25 dwellings and communal spaces arranged in blocks around a courtyard. Apex Acoustics initially undertook a traditional noise impact assessment for the site in support of a planning application. A soundscape approach was later adopted on realising the potential for a soundscape design process to improve outcomes as the stakeholders of the development are also its future residents. This was identified as a great opportunity to bring soundscape research out of the

1 seena.sajeev@apexacoustics.co.uk 2 rebecca.romeo.pitone@apexacoustics.co.uk 3 danuta.nabiałek@apexacoustics.co.uk 4 julien.batten@apexacoustics.co.uk 5 jack.harvie-clark@apexacoustics.co.uk

academic realm and test its potential in contributing to residential design. To the best of our knowledge this is the first time a soundscape approach has been used for a residential development.

The site is located close to the River Tyne, with Scotswood Road to the south and Atkinson Road to the west. Noise from industrial sites located to the south, east, and west are less prominent in comparison to the road traffic noise. The site location and architect’s visualisation of the development (viewed from the NE) are shown in Figure 1 .

Figure 1: Site location (left) and architects’ visualisation of the development (right) 2. DATA COLLECTION AND ANALYSIS

The idea for a soundscape assessment was initially presented to the residents in a focus group session, where their concerns and aspirations about the sound environment of their future homes were also discussed. Further data collection followed ISO 12913-2 [1] and was carried out in two parts – in-situ and virtual soundwalks. The in-situ soundwalks were carried out with the residents, and covered weekday daytime, late evening and weekend periods. Binaural measurements with point-of- view video were made during the in-situ soundwalks and other appropriate times, which were processed to create auralisations and visualisations [2] for the virtual soundwalks. The virtual soundwalks were conducted over Zoom meetings in two sessions, one involving the future residents and another with soundscape enthusiasts including academics, consultants, and members of the public. The latter was aimed at investigating the ecological validity of the technique and was advertised through social media platforms, attracting a global audience.

The in-situ soundwalks covered four locations on the site with varied soundscape and proposed uses. The soundwalk locations are shown in Figure 2 . Locations 1 and 3 represent residential block locations, while 2 and 4 represent external amenity/ garden areas of the development.

Figure 2: In-situ soundwalk locations, with the proposed building massing in white

The virtual soundwalk with the residents covered indoor sound environment scenarios at locations 1 and 3 with windows partially open and closed, and the courtyard sound environment. The virtual soundwalk with members of the public included the same indoor scenarios and outdoor scenarios for locations 1 and 3, with the exception of the courtyard scenario, which was included on the residents’ request. A 16 dB broadband attenuation was applied for the partially open window scenarios, while the closed window scenarios were calculated based on the proposed glazing performance in the traditional noise assessment report. Sound was mastered in all auralisations and room reverberation added to all internal scenarios. The virtual sound environment for the courtyard was based on the auralisation of the external sound environment at location 3, but with a 10 dB broadband attenuation, as determined from the Cadna/A model set up as part of the traditional noise assessment. As the virtual scenes were presented online through Zoom sessions, no calibration of each listener’s experience was possible.

Questionnaires following Method A of ISO 12913-2 were used to record responses from the participants of the soundwalks for the outdoor scenarios. The questionnaire methods of the standard are more appropriate for outdoor sound environments; the questions pertaining to the indoor sound environment were based on the 2-D measurement system proposed by Torresin et al [3]. The questionnaires for the in-situ soundwalk, virtual soundwalks with the residents and members of the public are available as open access, see Figure 8. The questionnaires included sections where the participants could record their responses as scale values, and additional sections to record comments. Informal discussions were encouraged following in-situ and virtual soundwalks, which provided further insight into the previous experiences and expectations of the residents. Questionnaire responses were analysed following ISO 12913-3 [4].

Acoustic and psychoacoustic parameters were also determined from the binaural measurements as required by ISO 12913-3. These parameters however did not provide much information, nor could they be linked to responses or comments and are therefore not included in this paper. These are available on request. 3. RESPONSES TO IN-SITU AND VIRTUAL SOUNDWALKS

Analysed responses of participants to the in-situ and virtual soundwalks are available upon request. A selection of results which revealed interesting aspects of the perceived sound environment are presented and discussed in this section.

The measured daytime noise levels at locations 1 and 3 were different, however the soundscape assessment as shown in Figure 3 indicates quite similar responses to perceived affective quality. In complete contrast, although the measured night-time noise levels at locations 1 and 3 were similar, the soundscape assessment results are significantly different, as shown in Figure 4 . This may be explained on the basis of the type and traffic pattern of the surrounding roads. Scotswood Road to the south of the site is a dual carriageway with a steady traffic flow and is located at approximately 90 m from the site. The measured average noise levels due to this road only differs by 5 dB between daytime and night-time periods. Traffic on Atkinson Road to the west of the site often labours up the steep hill on which the road runs, briefly stops for a mini roundabout to the immediate north-west of the site before setting off again. The measured average noise levels due to this road differs by 10 dB between daytime and night-time periods. Intermittent traffic flow on Atkinson Road is not taken into account in the measured average noise levels. The proximity of the road makes the intermittent traffic feel more intrusive. This is especially true during the night-time, when the background sound levels are lower.

Figure 3: Median values of response to perceived affective quality – In-situ soundwalk during

daytime at the locations of residential block elevations

Figure 4: Median values of response to perceived affective quality – In-situ soundwalk during

night-time at the locations of residential block elevations

The median responses to the perceived affective quality of the external amenity areas were also significantly different, as shown in Figure 5 . This may again be understood on the basis of the surrounding road traffic pattern. The external amenity area at location 2 is situated in line with Atkinson Road coming down the hill. To the listener, the sound of cars approaches directly, to become close and fast. These sounds are described as looming sounds, which according Marsh [5] are perceived as particularly alarming as they represent the greatest danger and threat to life. Distant or receding sounds represent less danger, which is the case for the sounds at location 4.

Figure 5: Median values of response to perceived affective quality – In-situ soundwalk during

daytime at the locations of external amenity areas

Positive responses to the soundscape at location 4 is also understood to be due to the presence of more natural sounds, especially birdsong, which was indicated in the comments and during the informal discussions as contributing to a positive sound environment. This is consistent with research [6]-[10], which indicate that the presence of natural sounds in combination with road traffic and industrial noise, or in the background, generally tend to reduce the perceived annoyance to a sound environment. In a study of outdoor urban soundscapes [11], sound environments dominated by natural sounds were perceived as ‘pleasant’, which agrees with the median responses to the perceived affective quality as location 4. The soundscape approach therefore revealed aspects of the sound environment at the site which were not apparent from the traditional noise assessment considering only the average decibel levels, and not the content or characteristics of the sound environment.

Responses and comments recorded by participants for different soundscape scenarios had notable differences; where a sound environment was considered appropriate by one person, it was also considered intolerable by another. Informal discussions revealed that non-acoustic personal factors such as the different sound environments of their current homes, and previous experience of living in proximity to road traffic noise of different extents, was a strong factor in the wide range of responses and comments.

In comparison to the in-situ soundwalk, the median responses to the perceived affective quality at location 1 and 3 were more negative for the virtual soundwalk with members of the public, as shown in Figure 6 . This might be indicative of the limitations of the virtualisation techniques in providing an accurate re-creation of the in-situ sound environment. Placing the participants in a survey context could also have affected the responses. Other studies comparing responses to in-situ and virtual scenes have also suggested that people can indicate a greater tolerance when they are in-situ compared to when they are listening to a virtual scene, as described by Yilmazer et al [12].

Figure 6: Median values of response to perceived affective quality – Virtual soundwalk with

members of the public, daytime scenario at the locations of residential block elevations

The indoor sound environments were considered “slightly” appropriate by the residents for the closed window scenario at location 1 and partially open and closed window scenarios at location 3, as shown in Figure 7 . The partially open window scenario at location 1 was considered “moderately" appropriate. The reason for the closed window scenarios receiving a lower or similar scale value in comparison to the partially open window scenario was understood from the participant comments as due to birdsong being cut out; they preferred open windows with more birdsong, even if it meant more traffic noise. The absence of typical indoor residential sounds such as televisions, kitchen appliances, and conversations in the virtualisations of the indoor sound environments also meant the participants were listening keenly to outdoor sounds, which does not necessarily portray real life situations. This is another reason that virtual scenes may illicit a different response from in-situ responses, if the virtual scene is incomplete. This raises more questions about how “realistic” virtual scenes could or should be curated.

Figure 7: Median values of response to appropriateness of indoor sound environment –

Virtual soundwalk with residents

4. FACILITATING CO-CREATION AND INCLUSIVITY

The soundwalk experience helped the residents recognise design improvements for the sound environment of their future homes. One of these included better enclosing the courtyard area to provide additional protection from intrusive road traffic noise, especially reducing the opening in the north-west corner, as location 1 was considered more annoying during the night-time. The inclusion of a courtyard was in itself a positive design measure, which although not specifically intended for the purpose, provide all residents with a quieter façade. However, reducing the courtyard openings to better block out road traffic noise could not have been anticipated without the soundscape assessment.

Almost all the residents considered birdsong as a positive element of the soundscape, and therefore in the discussion following the virtual soundwalk suggested planting in the courtyard to attract more birds and thereby enhance natural sounds. The soundscape assessment revealed the external amenity area at location 4 as distinctly positive in comparison to location 2. It was therefore suggested by the residents to use location 2 for louder activities, e.g., children’s play area, and allocate location 4 as an area for quiet and peaceful relaxation. These design responses could only be realised due to the soundscape assessment.

The soundscape approach therefore facilitated co-creation and inclusivity where the residents could contribute towards improvements to the sound environment of their future homes. This approach is a step towards facilitating an “agile participatory urban soundscape planning process” which has been identified by Xiao, Lavia, & Kang [13] to have the potential to identify acoustic issues, lessen the gap between soundscape theory and practice, and achieve socially and environmentally holistic outcomes.

Although the soundscape approach revealed many interesting aspects of the sound environment and facilitated design improvements, it did not address all the specific concerns of the residents. During the initial focus group session, the residents had expressed concerns on whether the dwellings built to Passivhaus standards would completely block out external noise and lead to hearing more neighbour noise. However, it was considered that this could be addressed at the detailed building acoustics design stage, and it was decided to focus on environmental sound issues for the soundscape assessment. Future soundscape assessments could benefit from setting design objectives at the beginning which include all the concerns of the stakeholders.

5. CHALLENGES IN USING A SOUNDSCAPE APPROACH

There are standards and guidance available to conduct a soundscape assessment, however, there is little guidance currently available on how the results of the assessment may be translated to soundscape design. The assessment helps understand the perceived affective quality and appropriateness of sound environments, but how can we improve these, and by how much? For instance, what design measures can be incorporated to change a sound environment from “moderately” appropriate to “perfectly” appropriate for its intended use. In cases where positive design improvements may be identified, the soundscape assessment would have to be repeated to confirm if the improvements have led to positive perceptions of the sound environment. A challenge in this situation would be to ensure the consistency of results while repeating assessments, as a different soundwalk situation with different participants, moderators, or weather conditions may result in different responses. It is anticipated that future standards will provide better guidance in translating soundscape assessment outcomes to design interventions. Post completion evaluations would also provide greater insight into the efficacy of the approach.

It is vital but difficult to ensure inclusivity. Median values of participant responses have been presented, following ISO 12913-3. Review of individual responses and comments indicate that the individual responses can vary greatly. More guidance and research are required to assist in

determining whether possible design interventions from a soundscape assessment should focus on the more sensitive or average group, or if an in-between is possible.

Timing of the soundscape assessment is a key aspect which would determine if the assessment outcomes and the resulting design interventions can be implemented in the design of the development. The soundscape assessment for Co-HUT was undertaken at a much later stage of the development, after it had been submitted for planning approval, such that major design changes were no longer feasible. A soundscape assessment could yield better design outcomes if implemented during the early stages of the development.

It might not always be feasible to include larger groups of people in in-situ soundwalks. Virtual soundwalks are a great alternative which can be more widely accessible. Allowing participants to take soundwalks at their own time and convenience would also remove the factors which affect responses when people are placed in a survey context. Such a virtual soundwalk for the CoHUT site is now available on our website . Further research is required to understand the limitations of the virtualisation techniques and to ensure ecological validity.

Another challenge is the resources, technology, and time required for a soundscape assessment in comparison to a traditional noise impact assessment. The resources required for the assessment are significant, and the potential extent of practical, positive outcomes cannot be determined. More case studies are required that can demonstrate how a development can be enhanced with the input of a soundscape approach.

6. CONCLUSIONS

Soundscape approach applied to a residential noise impact assessment revealed many interesting aspects indicative of both the scope and limitations of the approach. The soundscape assessment revealed aspects of the sound environment which were not evident from a traditional noise impact assessment and helped facilitate co-creation and inclusivity where the future residents of the development identified design improvements for the sound environment of their future homes. As an evolving and emerging field, soundscape exhibits great potential to improve sound environments, and thereby health and quality of life. Appropriate guidance and research are required to ensure assessment consistency and ecological validity of virtualisations, and a process for translating the outcomes of a soundscape assessment to effective design interventions. This case study has demonstrated the large array of new skills that are required to carry out a soundscape assessment, and the greater array of new tools and skills that are required to undertake effective soundscape design. This will help future practitioners understand how to scope and undertake a soundscape assessment.

7. ACKNOWLEDGEMENTS

We gratefully acknowledge the architects Mawson Kerr and the CoHUT group for their willingness and active participation in this study. We are also grateful to all the participants of the virtual soundwalk and our colleagues at Apex Acoustics for the valuable feedback.

8. REFERENCES

1. ISO/TS 12913-2:2018(en), ‘Acoustics — Soundscape — Part 2: Data collection and reporting

requirements’, 2018. 2. J Harvie-Clark, R Romeo Pitone, ‘Acoustic Perception EvALuation in Buildings - The APEAL

Method’, Proc. Institute of Acoustics , vol 42, pt 1, 2020. 3. S. Torresin et al., ‘Indoor soundscape assessment: A principal components model of acoustic

perception in residential buildings’, Build. Environ. , vol. 182, p. 107152, Aug. 2020. 4. ISO/TS 12913-3:2019(en), ‘Acoustics — Soundscape — Part 3: Data analysis’, 2019. 5. J. Marsh, ‘There's a Bad Noise on the Rise: Looming Sounds Produce Behavioural Attentional

Capture’, ICBEN , 2021. 6. S. Torresin, R. Albatici, F. Aletta, F. Babich, and J. Kang, ‘Assessment Methods and Factors

Determining Positive Indoor Soundscapes in Residential Buildings: A Systematic Review’, Sustainability , vol. 11, p. 5290, Sep. 2019. 7. W. Davies, N. Bruce, and J. Murphy, ‘Soundscape Reproduction and Synthesis’, Acta Acust.

United Acust ., vol. 100, Mar. 2014. 8. M. Romanowska, ‘The Warsaw soundscape - Structure and specifics’, Misc. Geogr. , vol. 18, Dec.

2014. 9. C.-J. Yu and J. Kang, ‘Soundscape in the sustainable living environment: A cross-cultural

comparison between the UK and Taiwan’, Sci. Total Environ. , vol. 482–483, pp. 501–509, Jun. 2014. 10. A. Gidlöf-Gunnarsson, E. Öhrström, M. Ögren, and T. Jerson, ‘Good sound environment in green

areas modify road-traffic noise annoyance at home’, pp. 1579–1587, 2009. 11. Ö. Axelsson, M. Nilsson, and B. Berglund, ‘A principal components model of soundscape

perception’, J. Acoust. Soc. Am ., vol. 128, pp. 2836–46, Nov. 2010. 12. Semiha Yılmazer, Zekiye Şahin, ‘A comparative study on soundscapes in real (RE) and virtual

(VE) open offıce environments.’, INTER-NOISE and NOISE-CON Congress and Conference Proceedings, InterNoise20, Seoul, Korea, pages 2990-3981, pp. 3235-3253(19) 13. J. Xiao, L. Lavia, and J. Kang, ‘Towards an agile participatory urban soundscape planning

framework’, 2018.

The introduction and links to the virtual

soundwalk is available here

Virtual Soundwalk Questionnaire