Information Technology Reference
In-Depth Information
reference item, no significant differences could be
found (T = 630.50 and T = 75, resp., p > 0.05, ns).
The data analysis showed that the ratings of
the tonal quality degradations in the active session
differed significantly from those in the passive
session. The low-pass filtering in the active session
was rated as being less perceptible compared to the
passive session, for which active players turned
into passive viewers. More generally speaking, the
experiment shows that an influence of interaction
performed in one modality (visual-haptic) upon
the perception of quality in another modality (in
this case, auditive) is possible. Thus, cross-modal
influences are possible.
In order for a cross-modal influence to exist,
the characteristics of stimuli and interaction/task
must be carefully balanced. At this time, it is not
possible to determine or quantify that balance a
priori. However, some of the influence factors
that contribute to this balance have been identified
in the salience model in Figure 2 above. These
influence factors need to be quantified and this
is a task for the future.
processes. Still, whether it is possible to come up
with a generalized model of cross-modal percep-
tual processing at all is highly questionable. It is
assumed by many that its complexity exceeds by
far the possibilities for designing a suitable model.
Yet, it seems feasible to aim at perceptual models
that are valid for certain perceptual scenarios only.
A specific game-playing scenario can be one of
these, as factors like setup (computer screen,
loudspeakers/headphones, input devices) and task
are of rather small variance across users, given
a certain use case. This has been demonstrated
in the game example above. A salience model
as described in this contribution could therefore
serve as a starting point for the exploitation of
salience effects.
Saliency is closely related to distribution of
attention and perceptual capacity limits. The
experimental results summarized in this chapter
indicate that effects of capacity limits are more
dominant inner-modally than cross-modally. At
the same time, capacity limits seem to be more
predictable inner-modally than cross-modally.
Unless we have better models of the percep-
tual processing underlying the generation of a
subjective quality impression, it will be difficult
to predict the perceived quality of audio in a multi-
modal context in general, or in a game context as
discussed here. Nevertheless, both the experiments
described, and the literature and effects reviewed
here, suggest that there is potential for exploitation
of such perceptual constraints.
Future research should therefore concentrate
on methodologies for the subjective evaluation
of audio-visual quality, or multi-modal quality
in general. Only a few recommendations exist
for performing audio-visual experiments and the
impact of interactivity—as naturally given in
any gameplay—on the perceived quality is, until
now, simply not considered at all. Once proper
recommendations exist, it will be much easier to
compare and validate experimental results, thus
paving the way for a quantification of the salience
model described in this chapter.
sUMMArY AND cONcLUsION
This chapter has reviewed some of the most
important issues of perceived quality of audio
in computer games. The main conclusion is that
audio quality in games, as perceived by a game
player, is not independent of other factors (apart
from sound quality itself). Because games usually
provide information and feedback to the player in
more than the auditory modality, it is necessary
also to take into account other modalities when
judging the impact and quality of audio. A rating
of audio quality alone, without the gameplay
context, is not meaningful.
The physical mechanisms of human auditory
and visual perception are well understood. Cross-
modal interaction between the two domains, that
is, perceptual processing in the human brain, needs
further research, before it is possible to model such
Search WWH ::




Custom Search