16 May 2021

Sound Visualisation in Religious studies? Some Preliminary Thoughts

Introduction

This work explores the applicability of utilising visualisations of long-duration field recordings within religious studies. It suggests that, while needing significant theoretical and methodological groundwork, visualisation can complement current interests in sound within the academic study of religion. Acoustic monitoring as a tool for animal conservation has been key in soundscape ecology, a field dedicated to the study of acoustic environments, or soundscapes [1]. A key component of this is field recording as an aid for monitoring changes in environmental health. Large acoustic “data” sets are impossible to listen to entirely, meaning that researches employ visualisations to elucidate any meaningful insights that such data may contain [6].

The Spectrogram

One such visualisation, and the visualisation utilised in this work, is the spectrogram (or sonogram), a graphical representation of sound (or any signal). Briefly, they are produced with a Short-Time Fourier Transformation (STFT), a common procedure in signal processing. On the visualisation, the horizontal dimension represents time, and the vertical dimension pitch or frequency in hertz (Hz). One reads the graph linearly from left to right (like notations in sheet music), with higher frequencies presented higher on the graph and vice versa. Studying spectrograms have provided insight about our acoustic environments. Bioacoustician Bernie Krause, for example, argues for the “niche hypothesis,” in which sounding organisms follow an internal organisation, like musical instruments in a orchestral arrangement, that can be deduced from a spectrogram [3]. The nature of the visualisation make them also particularly useful to investigate environmental change over time, as further discussed by Krause and Farina [4].

krause.jpg
Figure 1: “The Great Animal Orchestra” audio-visual installation by Bernie Krause and the United Visual Artists (2016). On the revolving spectrogram, annotations point to the frequency a particular animal is occupying (Courtesy of the Fondation Cartier pour l’art contemporain (2016), “https://www.fondationcartier.com/en/collection/artworks/the-great-animal-orchestra”).

The above suggests their application in the human domain as well, especially in questions of social change. In this sense, this work seeks to understand, for example, how trends of secularisation or other religious processes can me visualised with acoustic data. Changing religious landscapes have received much attention by scholars of religion, however, there is little research on how these changes affect their respective soundscapes. Studying soundscape data can point to new ways of understanding these processes, and help reveal research avenues normally hidden from conventional approaches.

Methodological Concerns

To better understand the above notions, the author carried out field recordings at the King’s College London Strand campus chapel, a Church of England site. These were recorded over a week period (during term time) between 3pm to 3:50pm each day. The chapel, located within the university campus itself and situated between social and teaching spaces, provides a unique case-study for the methodology outlined here. The location means that it is often utilised for non-religious functions, such as a place for students to meet friends or have their lunch. While arguably a good location, the choice was, unfortunately, poorly timed. It coincided not only with choir and organ practice, but also the University and College Union (UCU) 2020 strikes, and the advent of the current pandemic. These factors have, therefore, deemed the research here as tentative, with the intention of inviting further discussion about the general methodology presented here.

recording.jpg
Figure 2: The author’s recording setup in the Chapel.

Regarding Field Recording

During recording, the author took notes about what was happening in the chapel at a given time. Recordings themselves were done with a Zoom model H2n field-recorder at a standard 48 kHz/24 bit format. Ethical approval was granted for this project. Recordings were first processed by concatenating them to create a single sequence of around 7 hours. This was subsequently visualised with Sonic Visualiser, a program for viewing and analysing the contents of audio files1. This program was especially preferred over others due to its configurability, ability to include annotations, and its optimisation for use with large audio files. Please see the attached “png” file for visualisation here.

On Visualisation

This work understands that data visualisations should, on one hand, be visually accessible and engaging, and on the other, be interactive with an author-driven narrative [5]. The goal of developing a narrative of a “changing religious soundscape,” is to promote discussion and raise awareness of our relationship with our acoustic environment. In other words, this work seeks to make sense of complex phenomena (sound), and to simplify it to create a platform to tell a story. The nature of a spectrogram, being essentially a linear time-line of sound, makes it a particularly attractive form of this type of story-telling. Any visual narrative it may contain, is immediately apparent to the observer, in terms of the strong visual dynamics of sound there-in.

Caveats

Spectrograms bring forward problems in terms of visual attention, in that crucial information may be “hidden,” due to being of less amplitude, and therefore lighter [2]. This is evident in the perceived loudness of the organ, with other sounds drowned out by its visual noise. Furthermore, with subtle sounds, the Hermann grid effect is evident. Here the eye finds it difficulty to navigate absolute gradients between frequencies, rather than appreciating the actual relative difference between them. Further problems relate to the relative scarcity of spectrograms in public consciousness, and therefore the lack of vocabulary to efficiently interpret them. However, the purely visual effect of a spectrogram as a measure of sound quantity is easily observable, and (arguably) needs little training to appreciate. For example, a key narrative of Bernie Krause’s visualisation is the gradual disappearance of natural sounds in his spectrogram works. The impact of this hardly needs explanation when contextualised within the greater discussion of climate change.

To address some of the issues would involve further spacing out the spectrogram to contextualise its contents better. Some data that is too dynamic for sparking insights or making an impact (i.e. “chart junk”) [7], may need to be discarded. However, such actions may hamper the actual truthfulness and integrity of used data and the narrative it tries to portray.

Conclusion

This work sought to explore the idea of utilising long-duration acoustic recordings and visualisation within humanities research. While tentative, the basic framework presented here is hoped to be utilised for larger projects in the future.

References

[1] Almo Farina. Soundscape Ecology. Springer Netherlands, Dordrecht, 2014. [ bib | http ]
[2] C. G. Healey and J. T. Enns. Attention and Visual Memory in Visualization and Computer Graphics. IEEE Transactions on Visualization and Computer Graphics, 18(7):1170--1188, July 2012. [ bib | http ]
[3] Bernie Krause. The Great Animal Orchestra. Profile Books, Ltd, London, 2012. [ bib ]
[4] Bernie Krause and Almo Farina. Using ecoacoustic methods to survey the impacts of climate change on biodiversity. Biological Conservation, 195:245--254, March 2016. [ bib | http ]
[5] E Segel and J Heer. Narrative Visualization: Telling Stories with Data. IEEE Transactions on Visualization and Computer Graphics, 16(6):1139--1148, November 2010. [ bib | http ]
[6] Michael Towsey, Liang Zhang, Mark Cottman-Fields, Jason Wimmer, Jinglan Zhang, and Paul Roe. Visualization of Long-duration Acoustic Recordings of the Environment. Procedia Computer Science, 29:703--712, 2014. [ bib | http ]
[7] Edward R. Tufte. The Visual Display of Quantitative Information. Graphics Press, Cheshire, Conn, 2nd ed., 8th print edition, 2013. [ bib ]

Footnotes


rss | license