Physical Modelling for Sound Synthesis - Game Sound Technology and Player Interaction: Concepts and Developments - page 356

Some readers may have noticed that, in all
the studies described in this chapter, the focus
was rarely on formal listening tests. Instead, the
developers were usually the ones to evaluate the
sounds produced. Although anyone working in
this research area should have good listening skills
there is undoubtedly much to be learned from
carrying out extensive listening tests on a range
of people, including subjects with and without
experience in sound design, both gamers and non-
gamers. The results could reveal failings/overkill
in the current techniques and should inform future
research projects. This should, in turn, lead to
improvements in efficiency which is desirable in
any aspect of computer game programming, and
certainly in audio programming.
Finally, there are still barriers to overcome in
the adoption of physical modelling for sound syn-
thesis in common game engines. While Menzies
highlights the difficulties encountered in linking
sound synthesis to a physics engine (Menzies,
2007), he also stresses that these difficulties can
be overcome. So far, many projects have used
a physics engine by harnessing the information
already available from it, but no physics engine
has been adapted for sound synthesis from the
inside. The potential benefits of this are worth
exploring as, in a commercial sense, an in-built
sound synthesis module could give a physics
engine a competitive edge.
Looking forward to a time when the meth-
ods discussed here can be fully integrated into
computer games, it is important to view physical
modelling not as a standalone practice, but as a
branch of procedural audio. Within the context
of computer game sound effects there is a con-
tinuum from pure physical modelling to purely
abstract synthesis methods. Audio programmers
will require not only the knowledge but also the
creativity to know which part of this continuum
they should use in different situations in order
to create the most compelling sounds with the
resources available. Furthermore, game designers
will need to be aware of the new possibilities made

available by developments in physical modelling
and procedural audio.
At a minimum, games are set to become more
realistic due to improved audio. Each time a game
is played, a unique soundtrack, tailored to that
game experience, will be created. The subtleties
therein will match that which the gamer's intuition
expects based on what he or she sees and this will
lead to a more immersive experience.
Beyond this, new gaming possibilities are
opening up, ranging from small enhancements
of current common situations to completely new
prospects. For example, a common puzzle in The
Legend of Zelda: A Link to the Past (Nintendo,
1992) was to identify the weak point of a wall so
it could be destroyed. A weakness was usually
visible and also produced a different sound when
struck with a sword. However, the sound produced
by striking either a strong or weak point was
always one of two samples. With the incorpora-
tion of physical modelling into this situation, not
only could the sound of striking a wall contain
information on the material and thickness of the
wall, but also on the sword used to strike it. This
would still hold true even if the sword had been
created uniquely by the gamer during gameplay,
and this links to another range of possibilities
facilitated by physical modelling. Games will be
able to produce sound for objects that were not
conceived during development. As mentioned in
the introduction to this chapter, this is particularly
desirable in games such as LittleBigPlanet (Sony
Computer Entertainment Europe, 2008) which
encourage users to create their own unique con-
tent. Indeed, such games might require the user
to fashion objects that sound a particular way in
order to solve a puzzle. As also mentioned earlier,
with the increasing sophistication of physics in
games, we have now reached the point of objects
being realistically shattered, creating a potentially
unique combination of fragments each time. Again,
any shattered pieces will not have been conceived
during a game's development and so the only
way to create sound for them is by synthesis at