Cutting-edge data collection by citizen scientists maps a world of sonic diversity
Consider the following seemingly unconnected scenes: a slug eating a lettuce leaf, disposing of an unexploded bomb in Vietnam, and an office on an offshore platform in the North Sea. The chances are good that you imagined them visually, you ‘pictured’ them. We live in a strongly visual culture, where visual information is the norm and the environment is often described in visual terms. Yet, we know that there is much more to our sensory capability than just vision. Exploring people’s auditory appreciation of our environment is the subject of a pioneering project from the Acoustics Research Centre in the School of Computing, Science & Engineering.
The “sound map of Britain” project is an interactive examination of soundscapes and the construction of an online sound map. The apparently random examples mentioned earlier are just three of the 1,700 soundscapes that have been uploaded to the project. It has received high-profile coverage by participants from around the world including the BBC, Times2, The Daily Telegraph, The Guardian and local papers.
A soundscape is the aural equivalent of a visual landscape. It’s the sum of all sounds at a particular place and time. In some ways, it is more all-embracing than a landscape, because it represents 360 degrees of stimulus, not just the view ahead. You can turn away from a landscape, but a soundscape completely surrounds you.
When it started, the project was limited to just the Manchester area, but the widespread use of smartphones and the all-pervasive reach of the Internet meant that the researchers received submissions from all around the UK. Then, as online publicity took off, the project became global, with data coming in from around the world. This set the research apart from traditional sound data gathering, which used to be highly localised.
The project was careful not to be prescriptive about collecting a specific kind of sound because the aim was to gather as broad a range as possible. A soundscape could even be silence, although that is difficult to find outside an anechoic chamber.
One word that was avoided in any publicity inviting people to submit soundscapes was ‘noise’. If the project had been ‘noise’ oriented then it would have biased submissions towards negative sounds and environments. The term ‘soundscape’ is neutral; it reflects both negative and positive sound experiences.
There was an important objective beyond creating an interesting collection of soundscapes – a public engagement brief. The aim was to raise public awareness of soundscapes and the research potential around them. Participants not only captured the recording of the soundscape, but also rated the soundscape using different semantic scales: from good to bad, how exciting, eventful, tranquil and other descriptors. They were also invited to explain why they chose to record their particular soundscape.
One of the hallmarks of this project is the collection of data by smartphones – in this case Apple’s iPhone. In fact, iPhones proved to be so useful that the project in its current form would have been impossible just six or seven years ago.
As early as 2006, Salford researchers were looking at using smartphones to collect soundscapes. Relatively simple mobile phone technology had already been used to collect research data. Health workers had implemented schemes in countries without a landline infrastructure, using SMS texting to transmit data. However, the collection of soundscapes was a more complicated undertaking, which needed the power of more sophisticated smartphones. These have many advantages for soundscape collection: they are widely used, include a microphone, are powerful in terms of processing power and have the potential to record and upload research data.
This approach was tried; the predecessor of the current iPhone app was written for the previous generation of smartphones using a Java mobile edition. The problem was not a technical failure to operate, but the number of hoops users had to jump through to successfully send a soundscape. In those ‘early’ pre-iPhone days, the process of uploading directly from many phones was greatly hindered by a lack of bandwidth – people simply didn’t have good enough data contracts, so they had to go to a computer, plug in the phone, copy over the files and upload to the project’s website via their computer’s Internet connection. It was a real hassle, and the longwinded process killed the spontaneity of capturing a soundscape on the spur of the moment. It put off most people and only a few determined individuals sent in submissions.
In 2009, PhD student Charlie Mydlarz needed a new mobile and bought himself an iPhone. He realised that the simple interface and increased data allocation of iPhone subscribers meant all the technical and human interface pieces were now in place for an easy-to-use system that enabled users to upload soundscapes straight from their phone. A worrying lack of submitted data for his thesis was the final spur he needed to develop the iPhone app. Fortunately, it worked – before the iPhone and its app arrived there were about 300 to 400 submissions, but at last count there were over 1,700, thanks in no small way to the iPhone. To upload a soundscape, users now need an iPhone, the bespoke Salford recording app and a few taps on the screen.
Apple operates a closed system for its iPhone apps – all apps, even free ones, must be submitted to them for approval before they can become available for users to download. It can be a slow process, and the Salford app’s first submission took three weeks for approval. Any subsequent update, whether to iron out bugs or to simply add new functionality can take another two weeks for each new iteration of the system.
This was one of the few drawbacks with an otherwise vastly improved soundscape capture tool. If a fault develops, it can create real problems. After the iPhone app was launched, the project was covered by BBC click, which ran a story focusing on the app. Unfortunately, there was a system fault. Although it was a relatively small error, the effect was devastating for the project, because if an app is downloaded but doesn’t work first time then online feedback can be extremely negative, putting off potential new users.
Fortunately for the project, the medium for this feedback produced the ideal solution to bypassing system problems and communicating easily and directly with users – social media.
In terms of dissemination, social media played a big part in the growth and continued success of the project. It’s not enough to have a good data collection tool, even one that is easy to use, you need to be able to engage with and respond to your field ‘researchers’. The power and flexibility of social media meant that it also became a useful recruitment tool and helped to reduce user attrition if there were problems or glitches in the system.
The soundscape project uses all the usual social media outlets. There’s a Facebook page, which is active as a discussion board for people to debate sounds and talk about soundscapes. In the app itself, you can click a button so that when you upload the recorded soundscape, it will automatically post to Facebook and Twitter.
From a technical point of view, the social media aspects of the project are pretty much self-sustaining, requiring only occasional system management. However, the key to running a successful social media system, and ultimately a responsive and successful public engagement project, is you have to remain active. You have to strike up conversations and be very quick to respond to questions, suggestions for improvements and updates on any problems. Facebook became the official project discussion board, the social hub of the system. So even if there was an app problem that might take a couple of weeks to sort out, users could be reassured and kept in the loop over progress.
Twitter also proved its worth for fault finding. Users tweeted about the project: ‘this doesn’t seem to work’, or ‘how do I do this?’ and helpful replies often bounced back from other users in real time. Another useful aspect of social media was providing positive feedback. It ‘opened my ears’ was one quote, which helped to keep up the morale of hard-pressed researchers. It showed that users were starting to engage with the project and understand the difference between hearing and listening, really appreciating and thinking about the sounds around them.
Although it does not fit the conventional definition of ‘social media’, the sound map lets users listen to each other’s soundscapes and helps to reinforce the shared community aspect of the project. The interface is easy to use, and is one that most Internet users are familiar with – Google Maps.
Soundscape locations are precisely placed on the sound map, thanks to the app including the phone’s GPS data. For the occasional submissions that don’t use the smartphone and instead opt for upload via the website, then you still have to enter the location of the soundscape.
Despite the success of using the iPhone, it did limit data submissions to people who use the Apple device. This had two effects; the first was regular requests for a version of the soundscape app for people who don’t use iPhones, in particular a version that runs Google’s operating system – Android. Such enthusiasm from non-iPhone users was encouraging, but expanding the system would have doubled the software element. Sticking with just iPhones meant that data was submitted in a consistent manner, using essentially the same hardware and distribution platform. That said, the team is considering writing an Android version, but it’s fundamentally a question of time and funding.
The second effect was that just using iPhones meant users were a self-selecting demographic. The results were undoubtedly male biased and the age range that took part in the project closely mirrored that of smartphone use, roughly 24 to 34. It was essentially a trade-off between the widest possible range of participant versus maximising the number of people who can easily access the project. Groups who tended to miss out were the over 65s and the under tens; those who tend not to have smartphones.
One of the most successful aspects of the project was a key requirement of funding – public engagement and the concept of the ‘citizen scientist’. When the Salford team rebid for funding, they included a public engagement element to their proposal.
They got the funding from the EPSRC, as part of a Partnerships for Public Engagement (PPE) grant and the remit was to produce research whose main aim was to raise public awareness of the topic of soundscapes and the sound environment. Rather than talking about people who take part as ‘subjects’, the Salford team viewed them as participants – everyone who makes a submission is a part of the research team. In some ways it is a little like amateur astronomy, where anyone can make a valuable contribution. The data is ‘crowd-sourced’.
This has three important advantages.
First, it engages the public. People who might otherwise have had no opportunity to take part in a serious academic project at a leading research centre can make a valid contribution.
Second, submissions were received from all over the world; from places a small research team could never have the time or resources to reach: from Antarctica to northern Norway, and from Kabul to Burkina Faso and Paraguay, people have supplied soundscapes.
Unique access is not just a question of distance from Salford, but also privacy. One submission from the Mid-West in the States is a family watching television. You only get that kind of data from citizen scientists. The classic researcher, with a microphone and recorder would never get such intimate access to people’s lives and the soundscapes they inhabit.
Third, and perhaps most important, if a group of paid researchers are used then you will only get what they are interested in. By using the public, you get a vast range of submissions, from the mundane to the truly extraordinary.
Soundscape research is ultimately about people, so it makes sense for the collection of data to be human centred. The project worked because it’s about members of the public. You need to have the public playing a central role, enfranchising them with choice. Normally, academics would choose to focus narrowly or set the agenda but with the public taking part, they make the decisions.
Many of the early submissions were from schools, particularly ones that helped pilot the project. Collecting soundscapes fitted in nicely with their curricula because of the active element of going out and collecting data. Teachers bought into the project and gave their students iPads to map their school and surrounding area.
The project team also worked directly with local schools using ‘sound walks’. Participants walk around a pre-set route in silence and actively listen to their environment. There are stop points where the researchers record and write down what they’ve heard. The walks range from a 10-minute city stroll to a one-hour tour.
The work with secondary schools proved tough but useful. Some of the best beta testers were school students aged 14 to 16. If they didn’t like an aspect of the pilot software they had a real no-nonsense attitude in expressing their views on usability and functionality. Quite a few changes were based on the feedback from school pilots.
Not surprisingly, when people are at a site of relaxation or taking part in recreation then generally they rate the soundscape higher. One might assume that people like to be surrounded by natural sounds, birdsong and waterfalls, but the data and feedback from users of the sound map show that it seems that humans like to be around other people and prefer to be in places where there is human sound. It seems our natural disposition is to be social beings, even in our choice of environment.
Taking part in the project also helped people discover more about their sound environment. Participants tweeted or emailed that they would take a different route to work because of a more pleasing soundscape.
Mydlarz said: “It was fun watching the soundscapes come in; you never knew what was going to turn up next. The slug eating lettuce was good, totally unexpected and just bizarre. The bats in the basement of a palace in India are great. You can hear them chirp, since not all of their sounds are above the bandwidth of smartphone – I assume it’s just ‘bat chitchat’ and not echo-location. Even the mundane soundscapes can be fascinating; there’s a tattoo parlour in Toronto, with background chatter and the sound of a needle. Last, but not least, I enjoyed the Antarctic penguins, but then everyone likes a penguin.”
As the world gets busier and ever more mechanised, there is a perception that urban soundscapes are becoming increasingly degraded, particularly with the phenomenal rise in car ownership. In the study, the most common negative sound is car and traffic noise. While this has certainly been true over the past few decades, things may now be slowly changing. One of the major reasons for optimism is the advent of electric cars. As more and more electric cars displace the internal combustion engine, the overall rise in noise pollution should reverse. This will increase the dynamic range of the soundscape because the general background level will drop, allowing more pleasant sound to come to the fore.
However, it is not necessary to wait the decades that the electric car revolution will need to improve the urban soundscape; positive steps can be taken immediately. A good example where the sound environment has been factored into design is by Sheffield train station. A large fountain was built into the square in front of the station and helped to create an attractive public space where people meetand sit. This square is by main roads, but the addition of a positive soundscape helped to offset the negative impact of traffic noise.
In the longer term, as new built environments are constructed, the Salford research team wants to encourage the inclusion of soundscapes at all stages of urban design. Modern public spaces often look good, but the soundscape is generally not considered as much as it could or should be. The hope is that soundscape projects will start to generate and promote appreciation of the sound environment so that it can become a driver and a factor in urban design.
By engaging the public in soundscape research, you can get the public to comment, discuss and talk about soundscapes; thereby raising awareness. Research can also be used to influence policy, building standards and legislation, because it provides a firm body of evidence.
Subject to time and funding, the next stages of the soundscape project will look at more focused studies. One of the strengths, but also drawbacks, of the project so far is that submissions were very varied. Such a wide range of place types and sounds made analysis of significance difficult. The researchers are hoping to set ‘virtual fences’ around urban parks in major cities and focus on the soundscapes of sites of relaxation. This should give a more focused approach and tone down the amazing, but hard-to-analyse variety generated so far.
The virtual fences or ‘geo-fences’ let you trigger someone’s phone to alert them when they walk into a certain area. Again, the research will make heavy use of the power of smartphones to know their location thanks to in-built GPS technology. The study of soundscapes offers a new and fascinating perspective on our environment, both natural and manmade. The pioneering work at Salford has engaged a huge range of contributors, from local school students to a penguin spotter in Antarctica. It has made an important contribution to raising awareness of the opportunities that careful and considerate design can offer when building, inhabiting and using the world around us.
You can find out more about their activities at: www.acoustics.salford.ac.uk.
"While I was at school I became interested in audio equipment – designing and making small projects like radio station speakers. When the time came to choose a university course acoustics looked perfect for my interests, so I studied audio technology at Salford. The course covered the mathematics and physics of sounds, a lot of electroacoustic and transducer design, psychoacoustics (human response to sound); and even radio production.
My main strength was psychoacoustics and programming, which led neatly to the soundscape project and my PhD at Salford. My role started in 2007, and now I’ve just handed in my thesis. I’m hoping to get follow-on funding to do more soundscape work."