Posted by Tim Walton on , last updated
A lot of work that we do here at BBC R&D is related to object-based broadcasting. This is a technique that ultimately allows greater personalisation and optimisation of content by allowing the user’s device to piece together content in the best way for the individual, device and context.
As mobile devices become increasingly integrated into everyday life, increasingly people are consuming audio content in a wide variety of contexts. Whether you are on a train or in a café, your environment will likely influence how you experience audio content. One way to maximise your quality of experience when you are listening to audio in less than ideal environments could be to take advantage of the flexibility that object-based audio offers. For example, an audio mix could automatically adapt to the noise of your listening environment so that your experience is optimised. We recently carried out an experiment at BBC R&D to explore this topic.
We set out to investigate if environmental noise, such as the noise of a café or a train, influences the preferred mix of background and foreground audio objects. Foreground audio objects are those such as dialogue and key sound effects, whereas background audio objects are those such as background music and ambience. We used our state of the art listening room with 3D speaker setup to reproduce the environmental sounds you experience in a café or in a train. We then got 22 participants to adjust the background-foreground mix and overall listening level for various pieces of headphone audio content to their preferences. By comparing the mixes made with and without environmental noise we could see if environmental noise influences preferred background-foreground object mix.
So, what did we find? As well as the expected result that listeners increase the overall listening level in noisy situations, it appears that most listeners do indeed prefer slightly different mixes in different listening situations. Interestingly, the majority of listeners preferred mixes with a higher proportion of background objects in situations with loud environmental noise. This could be to mask the unwanted environmental noise with audio objects such as background music. However, this was not the case for all listeners; a smaller group did not change their mixes that much.
Results from this experiment illustrate that adapting an audio mix to the listening environment could be a useful technique to improve a user’s listening experience. It also highlights one area in which object-based audio could provide a greater experience to the audience. Over the next few months we intend to carry out more experiments to help us understand this further.
If you are interested in this work keep your eyes peeled for a BBC R&D white paper on this experiment, which shall be published in the coming months (update Feb 2017 - we have now published this paper). Additionally, this work will be presented at the AES 141st Convention in Los Angeles later in the year. For more information on this experiment, please contact me at email@example.com. This work was supported by an ESPRC industrial CASE studentship in partnership with Open Lab at Newcastle University and the BBC.
This post is part of the Immersive and Interactive Content section