The obligatory thesis word cloud about immersive audio

The word is out: I hereby present you my PhD thesis! After the final writeup in 2009, it took some time to get my viva arranged in August 2010. Sorry to keep you waiting…but now, it can be downloaded via this blog.

This thesis addresses the role of audio for immersion. Yep, the I-word…almost just as Impossible as it is Irresistible to designers. Yet, I’ve tried to come up with some useful theory for those who want to know more about the design of sound, music and speech in games and the conceptual decisions that are involved.

Let’s get things straight. This thesis does not:

• tell you which frequencies to amplify in order to make your music more immersive or explain you how to record immersive voice recordings (it’s not a cook book).
• provide you an answer to the question which psychological processes can be linked to immersion. There’s a lot of valuable literature available that investigates this subject.
• give you a final answer what to do when your game isn’t ‘immersive enough’. Game design is a discipline and you normally can’t fix game design flaws with audio!

Now what does it do? This work addresses the design decisions with respect to audio and immersion. I hope it offers audio designers, game designers and design researchers that are interested in the conceptualisation of game audio some illuminating insight into this matter.

Now go and create some Captivating Sound!

PhD award

ABSTRACT
This thesis explores the relation between game audio and (computer) game immersion. The main contribution in this research is a framework for the conceptual design of game audio in relation to immersion. A model for game audio is presented defining the communication by means of game audio during active gameplay. This model, named IEZA (Huiberts & van Tol, 2008), defines four conceptual domains of communication of game audio. IEZA was developed to enable a more detailed and consistent investigation into game audio and has been evaluated as a design resource in educational, academic and practical settings.

Immersion is considered to be one of the key factors making games worthy to play. The unity that seems to exist in the fact that many parties value this phenomenon is more apparent than real: there is a lot of dispute on the scope of immersion. Several classifications of immersion are compared for a better understanding of the nature of an immersive experience. The three‐dimensional SCI‐model by Ermi & Mäyrä (2005) is considered as valuable representation of the multi‐dimensional character of immersion.

Audio is an important constituent of most games and its role for immersion in games has hardly been investigated. In this thesis, audio is studied using the IEZA‐model and the SCI‐model, and several design issues are described. This yields a conceptual framework that describes various audio design issues that can be used to reflect upon conceptual decisions relevant for the design of audio in relation to immersion. This framework can be used for a conceptual analysis of design and is relevant for audio designers as well as game designers.

My PhD was a part-time PhD study at the University of Portsmouth in collaboration with the Utrecht School of the Arts.

Download the PDF here.

View the video files.

Huiberts, S., Captivating Sound: the Role of Audio for Immersion in Games. Doctoral Thesis. University of Portsmouth and Utrecht School of the Arts, Portsmouth, 2010.

A notification about your experience with this subject is warmly appreciated. Please send me a message using the contact form or use the comment form below.

I would like to express my sincere gratitude to all who supported the writing of this thesis: Jan IJzermans, Richard van Tol, Barbara Lotti, Tony Kalus, Adriaan Braat, Chris Creed, Han Onno ter Hoor, Micah Hrehovcsik, Rens Machielse, Paul Newland, Dan Pinchbeck, Dick Rijken, Hans Timmermans, Gerard van Wolferen, the other members of the Music Design Research Group, the gamers and moderators from the forums (Tweakers Forum, the 3Dgamers Forum, the Insidegamer Forum, the Gamers‐Forum, AudioGames Forum), the participants of PUGS, Espressofabriek and Hans and Riet Huiberts.

Game Audio Lab: an educational framework for research and design of realtime, nonlinear sound and music design. Photo © Sander Huiberts

The Music Design Group of the Utrecht School of the Arts under Jan IJzermans in the Netherlands has developed a lab for academic purposes which enables research and rapid prototyping of nonlinear audio. The system offers designers the ability to construct composite variables based on variables that are available in the original game system. In this paper, the architecture of the hardware and software as well as the motivation for this setup are discussed.

Nonlinear and adaptive systems for sound and music in games are gaining popularity due to their potential to enhance the game experience. The framework enables academics to map composite variables and adapt sound and music design in real-time during active game play. Composite variables were found to be more useful than mapping simple game parameters directly to the sound design. To give an example: mapping the state of an adaptive environment to the health level of the avatar often is too obvious for the player.

The paper also discusses the need for open game systems, which help designers to study sound design in realtime settings. Most game platforms are closed and protected by intellectual property. Interestingly, many academics mentioned this during the conference.

Unfortunately, I can’t publish the article here, so if you are interested it’s probably best to visit the AES website. The reference to the article is:

Huiberts, S. Van Tol, R and Went, K. (2009) GAME AUDIO LAB – AN ARCHITECTURAL FRAMEWORK FOR NONLINEAR AUDIO IN GAMES. In: Kelly, M. Proceedings AES 35th International Conference Audio for Games, 2009. Royal Academy of Engineering, London, UK.

Kees presenting at the AES. Photo: Sander Huiberts

Some video demos of applications (Half-Life 2 Source engine, OSC, Bridge software, FMOD and GStoolkit in Max/MSP) are listed below:

Credits for the game lab and these applications go to Arjen Schut, Mark Doeze, Maurice Alberts, Ramon Kerstens and Rogier Habraken. They truly did a wonderful job in this project.

Full demonstration Game Audio Lab Half Life 2: actual game play (select watch in high quality):

Simple demonstration: gunsound impact changes:

Simple demonstration: atmosphere dynamically changes with health status:

[For more information and the slides of this presentation, head over to adaptivemusicsystems.hku.nl]

To make things clear: we’re not talking about composing a song from the beginning to the end (linear music); the composer makes a large amount of musical ‘cells’ and the system selects new cells based on the rules of the composer (nonlinear music). Such an approach can be highly suitable for games, that mostly have a nonlinear character, as the music is able to correspond with the narrative or the presupposed experience of the player. And at least, we’re preventing the repetitive background track that drives players crazy.

The adeptive toolkit is still being developed, and now the developers have published a demonstration video of the toolkit with music composed by Than van Nispen. This is an interesting annotated video which shows the value of this system. What you see is an overview screen with music cells and a two-dimensional dipolar state diagram, which triggers the music. Notice the real-time “Notepad”-information which shows what is going on (real-time!).

The research focused on a basic nonlinear (adaptive) music structure, with a composer (non programmer) friendly approach to use any audio with a basic data design for nonlinear music, to enable prototyping and communication between game developers and composers.

[Watch the annotated nonlinear game music in Adeptive Toolkit v0.9 video here]

[Visit the Adaptive music systems Research group here]

[1] Vincent Akkermans, Bertus van Dalen, Siebe Domeijer, Than van Nispen tot Pannerden and Pınar Temiz.

Update: the adeptive toolkit has been published here. Read info from the devs below the cut:

The Adeptive Toolkit is an interface to a nonlinear music structure.

At the Utrecht School of Music and Technology, Bernard Herrmann’s flexible film composition methods were researched: a method of ‘cell composition’ – a cell being a small musical element or motif. Cells can be succeeded by different other cells, the choice depending on the demand of the (game/interactive) system (for example: more tension).

Cells are connected by transitions. The connections give the sense of a network, hence the term ‘cellnetwork’. Composers can chop their music into flexible cell-pieces, connect these with conditional transitions and value the cells with some variable, for example: “tension – 5″.

A flexible and responsive continuity in music can be created. A composer will then typically find himself making more and different transitions and ‘musical paths’ when he finds out how to serve the demands of the game better.

A video game engine communicates a request to the music structure, for example: avatar health 1. In this case, the music will transition to the ‘almost dead’ music IF the composer indeed provided a convenient transition network through his material.

Generally the ‘cellnetwork method’, reflected in the Toolkit, has been found an interesting way to deal with nonlinear music in contexts.The use of an open-source Max/MSP elaboration of the Cellnetwork Method could provide quick access to editing and expanding the system, as Max/MSP is used by many music technologists, sound designers and composers.

If you like to evaluate this toolkit, I’d suggest to contact the developers. They are interested in collaborating with other parties in order to expand this toolkit.