Bringing the tranquillity rating prediction tool (TRAPT) indoors – a case study at The Warren, Hull

James Oatley 1 , Mark Swale 2 , Clarke Roberts 3 Cirrus Research plc Acoustic House, Bridlington Rd, Hunmanby, Filey, North Yorkshire, YO14 0PH, United Kingdom

JJ Tatten 4 , Adam Pattrick 5 The Warren Queens Dock Chambers, 47-49 Queens Dock Ave, Hull, HU1 3DR, United Kingdom

ABSTRACT Tranquil environments are considered quiet, peaceful places where people can get away from everyday life, and for many years the University of Bradford’s Tranquillity Rating Prediction Tool (TRAPT) has been used to successfully predict the perceived tranquillity of outdoor spaces as a function of their auditory and visual modalities. The Warren, Hull, established 1983, is a citywide free project for marginalised and vulnerable young people aged 14-25 that helps to encourage and support them to take control of their lives and futures by engaging them in all of the decision-making that has impact upon the strategic and day to day running of project. Staff at The Warren have noticed that the young people often gather around a large public fountain to socialise and relax; the same fountain is due to undergo major work in 2022 and will be sectioned off from the general public for months. This paper addresses the challenges associated with bringing tranquillity indoors via considered and experimental application of the TRAPT to indoor spaces, and via empowerment of the young people to introduce tranquil auditory and visual changes to multiple spaces in The Warren. 1. INTRODUCTION

For many young people in Hull, The Warren is a place to come and feel empowered to make meaningful life decisions within a calm and open environment. The young people at The Warren thrive when they feel able to relax, learn, and be themselves without judgement of others; therefore, the young people at The Warren thrive with tranquillity.

Tranquil places are important for everyone, as they provide the chance to think, recuperate and unwind from the pressures of day-to-day life. Much research has already been conducted into tranquil spaces within smart cities, including the generation of a Tranquillity Rating Prediction Tool (TRAPT) from the University of Bradford, UK in 2010 [1]. The TRAPT may be used to predict the

1 james.oatley@cirrusresearch.com 2 mark.swale@cirrusresearch.com 3 clarke.roberts@cirrusresearch.com 4 jjtatten@thewarren.org 5 adam@thewarren.org

21-24 auGUST Scortsi event Cas La inter.noise | ee 2022

human feeling of tranquillity for any given outdoor space as a function of L Aeq , percentage of Natural and Contextual Features (NCFs) and any Moderating Factors (MFs). Equation 1: TRAPT equation [1].

𝑇𝑅= 9.68 + 0.041𝑁𝐶𝐹−0.146𝐿 𝐴𝑒𝑞 + 𝑀𝐹 (1)

Natural surroundings, stone walls, water structures, and birdsong as audible or visual features are typically quoted as having a quantifiable impact on the perceived tranquillity of outdoor spaces within existing papers that explore the validity of the TRAPT in urban and rural outdoor spaces. In fact, a major driving force behind this study was an observation of young people at The Warren gravitating towards a large water feature outside of the building in the city centre.

Inside, The Warren is a typical old town building, it has high levels of structural transmission of sound and reverberation down hallways. It consists of many function rooms including meeting spaces, music recording rooms, counselling rooms and a therapy room with two of the most popular being The Snack Bar, a café, and The Curve, a digital hub.

The young people have already been heavily involved within the design of the décor for The Snack Bar and The Curve. The aim of this study was to use these rooms to determine applicable NCFs and MFs for a communal space for young people, feed them into to the TRAPT, and validate the effectiveness of the TRAPT against the work of the young people to create more tranquil indoor spaces. 2. PART 1: INITIAL ASSESSMENT OF THE WARREN

The initial assessment of tranquillity at The Warren started in The Curve before moving on to The Snack Bar. The aim of the initial assessment was to solve the Equation 1 for the TRAPT by: • Determining the average perceived tranquillity felt by the young people on a scale of 1-10 • Identifying NCFs for indoor spaces at The Warren • Identifying MFs for indoor spaces at The Warren • Quantifying the impact of MFs on the perceived tranquillity and scaling the impact to solve Equation 1

The Curve and The Snack Bar were selected for assessment due to their highly functional nature and inherent popularity; all young people at The Warren experience being in these rooms.

The initial assessment involved 11 young people as active participants, who were asked to go about their daily business within the rooms awaiting a questionnaire. A Cirrus Research CR:172 sound level meter was set up on a tripod at head height, calibrated, and set to measure an L Aeq value while the assessment took place, and members of staff walked around the room guiding the participants through the questionnaire on perceived tranquillity.

As part of the questionnaire the young people were asked to observe and listen for 30 seconds and subsequently answer the following questions: 1. On a scale of 1-10 how would you rate the tranquillity in this room? 2. What audible features of the room do you like? 3. What visual features of the room do you like? 4. What audible features in the room do you not like? 5. What visual features in the room do they not like?

Responses to the questions were generalised into suitable feature categories and any features with only one point were discounted as outliers, as they did not affect a significant portion of the

group, defined by 2 or more mentions. The net positive or negative summations were calculated and grouped into MFs, NCFs or unassigned (man-made noises, such as talking, represented via the L Aeq ), as appropriate. Furthermore, images of the rooms were taken so that using the NCFs from the tables the percentages of NCFs could be determined for the TRAPT.

The method to calculated the percentage of NCFs was adapted from the 10 by 10 grid approach from “Tranquil spaces in a metropolitan area,” [2]. The grid was replaced by photo editing software and a freehand selection tool to separate the NCFs from the rest of the environment. 2.1. The Curve

Table 1: The Curve initial assessment of NCFs and MFs.

Feature Net Weighting (-11 to 11) NCF / MF Audible Visual Divisive

Talking 2 - x x

Loud talking / shouting -5 - x

Typing / clicking 2 MF x x

Décor 8 NCF x

Personal space / booths 3 MF x

Plants 2 NCF x

Multimedia systems 3 NCF x

Perspex screens -4 MF x

Figure 1: Images of The Curve, The Warren, Hull.

The mean perceived tranquillity of The Curve was found to be 4.82 with an L Aeq, 45 min of 72.6 dB. Décor, plants and multimedia systems formed part of the positive NCFs for The Curve, typing / clicking was taken as a MF due to its positive psychological impact, as were the personal spaces / booths. Perspex screens, installed to restrict Coronavirus, were taken as negative MFs, roughly synonymous with litter in a park within outdoor applications of the TRAPT, [2].

Approximately 58 % of the room was covered in positive NCFs, this value includes carpet as a positive NCF. 2.2. The Snack Bar

Table 2: The Snack Bar initial assessment of NCFs and MFs.

Feature Net Weighting (-10 to 10) NCF / MF Audible Visual Divisive

Music 5 MF x x

People too loud -4 - x

Table Tennis Noise 2 - x

General noise -4 - x

Loud music -2 - x

Décor 4 NCF x x

Friends 2 MF x

Table Tennis Table 2 NCF x

Figure 2: Images of The Snack Bar, The Warren, Hull.

The mean perceived tranquillity of The Snack Bar was found to be 4.10 with an L Aeq, 45 min of 75.2 dB. Music was found to have a positive impact and so was classified as a MF, and was the ability for the young people to see their friends across the room. Other noises were deemed suitably covered via the L Aeq .

Approximately 49 % of the room was covered in positive NCFs, this value excludes hard floor and hard furnishings. 2.3. Provisional TRAPT Moderating Factors Equation 1 was adapted so that the influence of MFs could be determined as per Equation 2: determination of MFs.

𝑀𝐹= 𝑇𝑅−9.68 −0.041𝑁𝐶𝐹+ 0.146𝐿 𝐴𝑒𝑞 (2)

The net influence of MFs on the perceived tranquillity was calculated as approximately 1.4 for The Curve and 3.4 for The Snack Bar. A linear mapping was applied to the MFs identified within Table 1 and Table 2 to generate suitable MF values for the TRAPT.

Table 3: Provisional calculated MFs for an indoor TRAPT, to the nearest ±0.25 .

Moderating Factor Adjustment

Typing / clicking 1.5

Personal space / booths 2

Perspex screens -2.75

Music 2.5

Friends 1

3. PART 2: SOUNDSCAPES AS TRAPT MODERATING FACTORS

Part 2 of this study involved engaging the young people to create two ambient soundscapes tailored towards improving the tranquillity of The Curve and The Snack Bar by applying their newfound knowledge of tranquil spaces. As part of this process the TRAPT, with its newly determined indoor MFs, was applied using music as a MF for the musical elements of the soundscapes in attempt to predict the impact of their proposed changes and understand how different types of musical sound may impact the MF.

In order to create the soundscapes, the young people were split into two groups, one group was assigned The Curve and the other The Snack Bar. 3.1. The Curve The group assigned to The Curve considered how the space may be made more tranquil while a large gaming / VR session was taking place. This influenced their decisions within the creation of the soundscape by helping them to decide exactly which sounds were tranquil and which were not.

The sounds which were deemed non-tranquil for The Curve included talking and shouting as such sounds posed as a distraction to normal working practices; The Curve is not always used as a gaming space. The sounds which were deemed tranquil included typing, rustling paper and a

mediative mindfulness-style soundscape, many of which the young people discover and listen to using YouTube and headphones while working.

The group were asked by staff at The Warren to consider aspects of the composition of a custom soundscape for The Curve and provided the following guidelines: • “Anything too interesting will distract me” • “I like it when it feels like I’m in a big space” • “Oh no that’s too scary” – in reference to minor keys • “It makes me feel warm” – in reference to a low sustained bass synthesiser

With this considered, a single chord played by wide-panned synthesisers with high levels of artificial reverberation and delay were utilised to keep the soundscape simple, big and non- intrusive. A secondary independent track containing typing, clicking and rustling paper noises was also composed to provide the remainder of the sounds deemed tranquil.

The audio equipment was configured to provide the most natural distribution of sound throughout The Curve. Two hi-fi speakers, panned far-left and far-right, were used to play the secondary track while a large stereo TV soundbar played the main musical soundscape, as presented in Figure 3.

Figure 3: Speaker layout for soundscape playback in The Curve.

Questions from the part 1 assessment were once again posed to the wider group while the soundscape played and a calibrated Cirrus Research CR:172 sound level meter recorded an L Aeq value at head height. The overall L Aeq throughout the duration of this period was 64.9 dB.

The following MFs were selected: • Typing / clicking • Personal space / booths • Perspex screens • Music and combined with a 58 % NCF value and the recorded L Aeq to generate a predicted tranquillity rating of 5.8.

The average tranquillity reported by the group was 7.2 providing a net difference of +1.4 between the predicted and actual tranquillity ratings.

The ambient soundscape was non-divisive, it was received by the group positively. However, the sounds of typing / clicking were divisive, as per Table 1.

3.2. The Snack Bar The group assigned to The Snack Bar found themselves considering the design of the soundscape during a time that they described as “eerily” quiet. Therefore, the group decided that the soundscape should include an element of chatting. This acoustically contextual element of chatting was not identified as a MF within the part 1 assessment and formed an interesting component of the wider study.

The group decided that they would not like to add any sudden loud noises such as passing busses, people shouting, or babies crying, to the soundscape. Much discussion took place on the subject of music which concluded with them wanting to create two soundscapes, one using the widely enjoyed “lo-fi beats” playlist from YouTube, and the other using a more diverse mix of their favourite songs [3].

The sound of people chatting was recoded by staff at The Warren on a busy day in The Snack Bar. Originally it was intended to use a generic café chatting sound loop from the internet, but the young people found the loop unsettling and preferred it when their friends could be identified within the sound mix.

The audio setup utilised six speakers to ensure that the sound-field did not sound “fake” or “lifeless.” Ambient chatter was played through two stereo PA speakers, and the music was played through 4 infill speakers, as per Figure 4.

Questions from the part 1 assessment were again posed to the wider group while each soundscape played and a calibrated Cirrus Research CR:172 sound level meter recorded the L Aeq values at head height. The overall L Aeq throughout the duration of this period was 72 dB for the lo- fi-based soundscape and 75.4 dB for the playlist-based soundscape.

The following MFs were selected: • Friends • Music and combined with a 49 % NCF value and the recorded L Aeq ’s to generate a predicted tranquillity rating of 4.7 for the lo-fi-based soundscape scenario and 4.2 for the playlist-based soundscape scenario providing respective net differences of +1.1 and +0.1 between the predicted and actual tranquillity ratings.

The lo-fi-based soundscape was non-divisive, it was received by the group positively, but the playlist-based soundscape was not deemed tranquil by two members of the group.

Figure 4: Speaker layout for soundscape playback at The Snack Bar.

4. PART 3: APPLICATION OF THE TRAPT USING VISUAL ELEMENTS

The final assessment involved the introduction of visually stimulating, and contextually acceptable elements to the rooms. A changing fractal video was introduced to the TV within The Curve and a projector was setup in The Snack Bar to display city buildings at night. All blinds were closed in an attempt to enhance the visual experience.

Figure 5 – City buildings in The Snack Bar (top left) and Fractals in The Curve (bottom right).

The average tranquillity rating for The Curve with the fractal video playing alongside the soundtrack was 8.6, a difference of +1.4 over just the soundscape. Likewise, for The Snack Bar the new average tranquillity rating became 6.1, a difference of +0.3 over just the soundscape.

The addition of these visual elements was regarded as tranquil by all participants. 4.1. Revised TRAPT Moderating Factors Additional MFs were created using the results from the part 2 and part 3 assessments so that the TRAPT would accurately match the average perceived tranquillity ratings by participants of the study in The Curve and The Snack Bar.

The only modification from the MFs in part 1 was to appropriately rename “music” to “radio music” to allow the different types of music to be distinguished.

Table 4: Calculated MFs for an indoor TRAPT, to the nearest ±0.25 .

Moderating Factor Adjustment

Typing / clicking 1.5

Personal space / booths 2

Perspex screens -2.75

Radio music* 2.5

Friends 1

Favourites playlist* 2.5

The Curve soundscape* 4

Lo-fi beats* 3.5

Fractal visuals + low light 1.5

City buildings visuals + low light 0.25

* Mutually exclusive moderating factors

5. DISCUSSION

5.1. Part 1 The assessment in part 1 revealed somewhat obvious factors such as the presence of friends, music and lack of Perspex screens leading to a more tranquil space. However, it also revealed a degree of comfort found in the sound of office-esq work and the benefit to the young people of having guaranteed personal space. The décor of The Warren, coincidentally designed by the young people, was generally seen as a highly positive NCF for the TRAPT alongside plants, multimedia systems, and the presence of a table tennis table.

Divisive elements of the environment included talking, typing / clicking, music and décor. The assessment revealed that divisive elements can become lost in application of the TRAPT. 5.2. Part 2 The assessment in part 2 revealed the positive tranquil impact that appropriate musical soundscapes can have on an indoor space. The custom-designed soundscape, led by the young people, increased the tranquillity of The Curve by +4, rendering it the most significant positive MF within the study. Furthermore, the introduction of lo-fi beats within The Snack Bar provided a significant +1 difference to the tranquillity rating when used in place of the radio.

On the other hand, a playlist containing the young people’s favourite music did not provide any significant difference in comparison to the radio. In fact, for young people not involved within this study it could be predicted that the marginal positive impact of this playlist would be lost.

Most interestingly, the assessment raised the concept of contextual appropriateness for MFs. The Snack Bar was described as eerie without the normal chatting; however, the chatting was undesirable within the design of the soundscape for The Curve. 5.3. Part 3 The assessment in part 3 revealed the undeniably positive impact that appropriate visuals can have upon the perceived tranquillity of an environment. It is particularly notable that the TV within The Curve was present and switched off throughout the entire study until this final assessment, when the closed blinds and fractal visuals resulted in a tranquillity rating +1.5 higher than without the visuals. Furthermore, the introduction of the city visual and closed blinds within The Snack Bar increased the perception of tranquillity by 0.25.

6. CONCLUSIONS

Each part of this study has explored new and significant MFs and NCFs for the TRAPT, allowing it to replicate the average perceived tranquillity felt by a group of young people from The Warren in The Curve and The Snack Bar. It has been found possible to use the TRAPT indoors in such a way, and the study has provided the framework for further investigations into the applicability of the new MFs and NCFs with the TRAPT to other rooms within youth centres.

The positive impact of non-intrusive audio-visual elements to select indoor spaces in a leading youth centre has been quantified using the TRAPT as a guiding aid.

It was found that use of the TRAPT must be carefully considered for two reasons: MFs, such as typing noises, can be generally positive but introduction of them can significantly lower the tranquillity for a minority; and that the appropriateness of MFs for any particular environment can change depending upon the general expectations of the environment.

7. ACKNOWLEDGEMENTS

We, Cirrus Research plc., gratefully acknowledge the young people and staff of The Warren, Hull, for working alongside us as part of this unique study using the TRAPT indoors.

8. REFERENCES

1. Pheasant, R. J., Watts, G. R, Horoshenkov, K. V., Tranquillity rating prediction tool (TRAPT).

Acoustics Bulletin , 35(6) , November 2010. 2. Pheasant, R. J., Watts, G. R, Horoshenkov, K. V., Tranquil spaces in a metropolitan area.

Proceedings of 20 th International Congress on Acoustics, ICA 2010, Sydney, Australia, 23 August – 27 August 2010. 3. Lofi Girl, lofi hip hop radio - beats to relax/study to , https://www.youtube.com/watch?v=5qap5aO4i9A .