Small Grants / PhD RA Project Showcase

On 3 October we held a showcase for the recipients of our Small Grants and PhD RA awardees at Edinburgh University and Edinburgh Napier University. Eight projects presented their progress so far and plans for future research.

Platform to platform: an investigation into audience engagement with digitised archives and their transformative impact across different online formats, Bruce Ryan and Hazel Hall

Hazell Hall presenting ‘Platform to Platform,’ photo courtesy of Bruce Ryan.

This project took diary entries by one-time Stirling resident Lorna Lloyd discussing her experiences living through World War II and transformed it into a narrative non-fiction podcast series. The research centres on the creation of a multi-faceted sound archive and evaluation of audience engagement with it. Listen to the podcast on the Malvern Museum website, and read more about what the researchers learned in their final project report.

Dynamic Range: immersive screenwriting for sound, James Mavor

James Mavor presenting ‘Dynamic Range,’ photo courtesy of Bruce Ryan.

Dynamic Range explores the dramatic potential of Augmented Reality (AR) sound in a mainstream television drama series using Dolby Atmos. When we stay close to the main character and hear what she hears, what does this sound like? How does the screenwriter script this on the page? And what story possibilities are generated by the data streams accessible through developing AR earbud technology? James presented a short video clip of a propposed television series making use of the technology as a central part of the dramatic narrative.

SOVRA: Subjective orientation in VR audio, Ethan Robson

Ethan Robson presenting SOVRA, photo courtesy of Bruce Ryan.

This project establishes a method for capturing listeners’ experiences of spatial audio in VR. This project moves beyond studying physical anatomy to the listeners’ subjective experiences of hearing in VR environments. There are many factors that affect auditory spatial perception, from levels of listening, reproduction hardware, noise floor, presbycusis, general health, cognitive load, levels of distraction, type of content and even the choice of task, amongst many others. Having the ability to identify where listeners perceive a sound emanating from can also utilised for diagnosing hearing impairments as well as head injuries.

Digital Munya 2.0, Glaire Anderson

Glaire Anderson and Deniz Vural presenting ‘Digital Munya 2.0’, photo courtesy of Bruce Ryan.

Digital Munya 2.0 is a research project developed by Islamic visual culture researcher Dr. Glaire Anderson on using the Unity video game engine to create an immersive digital experience of al-Rummaniyya, the only villa (Arabic munya) with extant remains from 10th century Islamic Spain. This and other villas of the time were a key feature of the landscape of the early medieval cities of Islamic Spain (Arabic al-Andalus). As these villas no longer survive as standing monuments, the Digital Munya experience gives the public a new way to understand important aspects of courtly life in medieval Islamic spain. With PhD RAs Deniz Vural (History of Islamic Art) and Dara Etefaghi (video game development and audiovisual art), the project updated the original Digital Munya build in Unity to further develop the immersive experience of the space by adding 3D models of artefacts and ‘lore’ about the artefacts and the villa, and to introduce auditory elements to convey the villa’s function as a medieval space of musical and poetic performance. Read more about Digital Munya 2.0 in the final project report.

Investigative Computing: an aesthetic approach, Martin Disley

Martin Disley presenting ‘Investigative Computing: an Aesthetic approach,’ photo courtesy of Bruce Ryan.

This project uses aesthetic methods to interrogate the essentialist assumptions of inherent value in the construction of machine learning models, specifically facial recognition technologies. Whilst the development of facial recognition models have been underscored by extensive critical work on their inherent fallibility, models centred on the voice remain an understudied area of AI ethics research despite falling foul of the same essentialising impulse. Through the production of an experimental short film, this project seeks to open up the social and technical limitations of these models for a general audience. In doing so, the project collaborators hope to articulate the potential of investigative computing to be used as part of creative practice to disseminate technical knowledge to non-technical researchers or to enable non-technical researchers to explore technical systems.

Tidesong: exploring the potential of online exhibition space through the creation of a data-driven, web-based art installation, Victoria Evans

Victoria Evans presenting ‘Tidesong,’ photo courtesy of Bruce Ryan.

Tidesong.app is an online evolving interactive artwork that allows users to make music with the tidal data from their local coastline, generating time and location specific audience experiences via a website and mobile app. This practice-based research project aims to connect audiences to the natural rhythms present in tidal landscape and asks to what extent an online installation can offer an embodied and situated art experience. Through a process of data sonification (the use of non-speech audio to perceptualise information) tidesong allows participants to create a musical track unique to them. The track is then sharable with other audience members, creating an evolving audio-visual artwork that maps the coastal environment. Read more about Tidesong in the final project report, and check out the artwork at tidesong.app.

Music Streaming as Global Cultural Diffusion, Tod van Gunten

Tod Van Gunten presenting ‘Music Streaming as Global Cultural Diffusion,’ photo courtesy of Bruce Ryan.

A feature of many cultural marketplaces is that success is very unevenly distributed. The most successful songs, books, films and other cultural products are wildly successful, while the majority of products have much more limited success. Sociologists and economists have studied the social processes producing superstar effects, as well as the patterns of hit-making on radio, but previous research has not studied these processes at global scale. Using a large dataset of Spotify information, the researchers were able to conduct analysis on many thousands of songs while considering different definitions of what constitutes “success.” They find, for example, that songs reaching the top 40 in 50 or more countries nearly always do so extremely rapidly: when global success comes, it usually comes very quickly. However, “medium” size hits follows a more diverse set of trajectories, implying that there may be paths to global success that do not require large, coordinated marketing campaigns. Read more about the first phase of this research in the final project report, and stay tuned for future research results on “medium” global hits.

Collaborative Creative Co-writing with an AI Author, Arabella Sinclair and Christopher Lucas

Arabella Sinclair and Christopher Lucas presenting ‘Collaborative Creative Co-writing with an AI Author,’ photo courtesy of Bruce Ryan.

Storytelling is an area where humans excel, using creativity and communication skills to pass on histories, introspect, entertain, and connect with others. This project investigates whether and to what extent Neural Language Models can be useful collaborative tools for writers creating short stories. As part of this investigation, the researchers are interested in exploiting the ability of NLMs to be fine-tuned to examples of stylistically specific language, and in creating several potential style specific AIs with which our human authors can collaboratively create with. Read more about the first phase of this research in this project report and stay tuned for the results of our ongoing experiments. If co-writing with an AI author sounds interesting to you please reach out to Arabella and Chris if you would like to participate as an author in their study.