| Summary: | This thesis analyses the benefits of using object-based audio as a production and delivery format in order to enable new audience experiences. This is achieved though a series of case studies, each focusing on a different user experience enabled by the use of object-based audio. Each study considers the impact of using object-based audio on the creative process, production workflow and audience experience. The first study analyses the audience’s use of the ability to personalise the mix of a live football match. It demonstrates that there was not a single audio mix favoured by all, and the ability to change the mix was valued by the audience. While listeners did adjust the mix initially, they tended to leave it at that setting and did not interact much once they made their initial selection. While there were three favoured mixes, over 50% of listeners did not choose one of these three mixes, indicating that only offering three options would not satisfy everyone. Modes of listening model the ways listeners deconstruct complex sound scenes into foreground and background categories ascribing different salience to foreground and background sounds. The second study uses this model to inform a series of card sorting exercises which result in similar foreground and background categories. However, rather than being unimportant, background sounds were present to convey ancillary information or to affect emotional responses and foreground sounds to expose plot or story events. This study demonstrated that this grouping was a meaningful categorisation for broadcast sound and evaluated how beneficial allowing different foreground and background audio mixes would be for audiences. It contains analysis of audio objects in the context of foreground and background sounds based on the opinions of the content creators. It also includes subjective testing of audience preferences for different mixes of foreground verses background audio levels across five different genres and four different loudspeaker layouts. It shows that there is no clustering of listeners based on their preference of foreground vs background balances. It also shows that there is significant variation of foreground and background balance preference between loudspeaker layouts. The final study goes beyond tailoring audio levels, balances and loudspeaker layouts and analyses the benefit to audiences of being able to adapt the story of a drama in order to set it in a location that is familiar to the listener. It shows that being able to set a radio drama in the location where the listening is taking place improves audience’s enjoyment of the programme. 75% of listeners who experienced the tailored version of the drama reported liking the story, compared with 65% of listeners who experienced a non-tailored version. The three studies also analyse the impact of object-based content creation on production workflows by documenting the challenges faced and discussing possible solutions. For example, providing writers with constraints when they are designing dynamic content and allowing sound designers time to develop trust in the technology when mixing content for multiple loudspeaker layouts. The original contribution to knowledge is to establish a new listening model applicable to constructed and designed sound experiences based on functional analysis of audio objects. This work also establishes, for the first time, a framework for the definition of an audio object based on the creator’s intended range of audience experiences. In addition the thesis also provides insights into how audiences interact with object-based content experiences and insights about audience attitudes towards using personal data to personalise object-based content experiences. Each study addresses the potential advantages of delivering object-based audio, assess any impact on the quality of the audience’s experience and analyses the challenges faced by production in the creation of these new experiences.
|
|---|
