To find out where things stand – and where we’re headed – I invited sound designer and game audio advocate Stephan Schütze to do a guest post to give you an overview. Here is Stephan’s post:
Game audio has come a long way in recent years. Its profile, tool sets and outlook are stronger than ever. Game audiences have high expectations of the audio that accompanies their favourite games and developers are investing more time and resources to audio production. This is a perfect time to take stock of exactly where game audio is currently at and consider some of the possibilities for the future in how we develop audio content across the many platforms we play games on.
Middleware Solutions
The term ‘middleware’ essentially refers to software solutions for game process management and asset implementation. There are various middleware applications that deal with audio, but I would consider the four main game audio tools to be (in alphabetical order):
• CRI ADX2
• Fabric
• FMOD
• WWise
Each application has its own methodology and feature set and choosing the best solution depends a lot on the needs of a project and the personal preferences of the audio team. It is safe to say, however, that the sophistication of the available tools has developed dramatically over recent years. All four of these applications have supported multiple significant titles across a wide range of platforms.
While the individual choice of which toolset best suits a particular project is a more individual one, the overall question of “why use middleware?” is still a common one. This question is not often asked by audio teams, but more usually by their development leads or studio heads, who require confirmation that the time, effort and expense of using an audio tool set will be advantageous to their project.
The game audio industry is still often asked this question, and I have a very simple and very direct response to the question.
Why should we use audio middleware?
• Your game will sound better
• Your game will use less resources
• Your game will require less programmer time to achieve equivalent results with your audio
Just to spell it out, that last point means using audio middleware will also save you money.
My personal opinion on this is that any studio that considers itself to be a serious developer of interactive material should be using audio middleware
My personal opinion on this is that any studio that considers itself to be a serious developer of interactive material should be using audio middleware in the same way they should be using source control software, debugging tools and all the other advances in development tools that are now considered essential.
Games are dynamic; so is game audio
Games are different to film and TV. I have said this so many times in articles, at conferences, in training and to students. The non-linear domain in which games reside means they are created in very different ways to film and TV. Non-linear media is experienced in very different ways to linear media. Game audio still lags behind in some aspects of non-linear development. This means, we have room for some great improvements.
Generative and dynamic audio is so much more than just cueing the music to respond events within the game. The toolsets available to audio teams have the power and control to create incredibly detailed and dynamic audio material.
Dynamic Environments
Game environments can be created from the smallest of audio assets that trigger with defined behaviour to fill a region of a 3D world. This can provide vertical, horizontal or even spherical depth of field. As the player moves through an environment they pass through layers that blend together and react to the player, other environmental factors as well as day/night and seasonal cycles.
A game audio environment is not made from a single recording of a forest or a jungle; it is built from the individual elements that would exist in that jungle. An insect can be positioned individually in 3D space and can be programmed to respond to the player’s proximity just as a cricket in real life will fall silent if it detects movement nearby. Birdsong is generated in real time to create a unique song every time it is heard, that song can alter to a birds warning calls if it detects a threat in its territory and ultimately resolve with the sound of wings as the bird flies off.
Music
Large orchestral scores with even larger budgets are a more common feature of AAA game projects. Equally, music generated in real time, controlled by properties that define the behaviour of music over time in relation to events and in response to player actions, are becoming powerful tools for narrative support.
There is a secret about these two approaches to game music that many people do not seem to have realized
There is a secret about these two approaches to game music that many people do not seem to have realized: The two methods are NOT mutually exclusive!
Audio teams seem to choose on method or another. Live musicians with strong thematic material or generative musical structures crafted and implemented carefully to produce a dynamic score during gameplay. I would argue that the best possible world is a combination of the two forms. Dramatic thematic material that accompanies significant events within a game, that underscores cut scenes and defines our wonderful characters AND evocative generative sound/music ambiences that accompany the many hours of exploration and highlight the underlying emotional content of an environment or expand on the threats that may exist in the shadows.
Sound Design
Even our sound effects can be created in dynamic ways that utilise the available assets to maximise on resources as well as sonic impact. Each sound file that we add to a project can become a building block to be used again and again across multiple sound events. This gives us incredibly efficient resource usage on all platforms. It also offers the opportunity for an explosion to be subtly different each time it is triggered, or footsteps that sound organic when implemented.
How do we do this?
For some people these ideas may sound challenging at best, unachievable at worst, but the technology to utilize many of these production techniques has existed for some years. What we need to be doing is educating our fellow developers and demonstrating the possibilities. The incredible potential for game audio is already being demonstrated by some teams, we need to realize across the industry that this is something we can all be doing if the desire exists and the determination is applied.
There needs to be a shift in thinking to understand that outstanding audio is not just reserved for AAA games
I have spent a lot of time over the years investigating three of the four middleware solutions I listed (Fabric, FMOD and Wwise) and to my knowledge they are all capable of far more than many audio teams realize. There needs to be a shift in thinking to understand that outstanding audio is not just reserved for AAA games.
Some recent independent games have clearly illustrated just how much you can achieve. Limbo, Braid, Machinarium, The Stanley Parable are all examples of small teams achieving incredible audio results.
I think we all need to be multi-skilled to work in game audio. Where film and TV often have a single specialist for each role, game audio is better served if we at least have a strong understanding across all aspects of audio production. Location recording can make you a better sound designer, understanding sound, music and dialogue processes will ultimately make you a better mixer. Even developing an appreciation of how sound design is implemented can make the creation of a sympathetic musical score more achievable. Above all else, passion and patience are critical, and a good set of ears is a big advantage.
The Future
HRTF, Dolby Atmos, procedural audio design: these are all ‘new’ areas of game audio that are still somewhat on the edges of our radars. Often we are just struggling to get all the audio into a project in the time we have. What formats, features and functions become more common in the future is, however, up to us to decide. An audience cannot appreciate a new format if we do not explore it and make the most of its potential. All the middleware developers will continue to advance their toolsets and functionality to allow the audio teams to achieve greater results.
How we use our time is important. Dedicating even a small portion of time to test and assess new tools allows us to glimpse potential futures and be inspired to attempt new things. The nature of our creative work means that many of us will constantly work towards improving our art form for our own satisfaction and for the enjoyment of our audience.
For new technologies such as the Oculus Rift and Project Morpheus to be truly successful, they MUST have audio that supports them.
The future of game audio may be interesting, but the present is amazing!
Those devices will succeed or fail based on how the audience responds to the experience and the audio will be a critical aspect of that success or failure.
The future of game audio may be interesting, but the present is amazing! There is so much potential in what we have right now that we just need to embrace a few scary new concepts and dive in as deeply as possible to really benefit from how the technology can support us in creating truly unique and engaging audio experiences within our game projects.