Information Technology Reference
In-Depth Information
the dimensions revealed by the experience and position the sounds of the corpus
occupy. The process of analysis is generally completed by relating perceptive
dimensions with acoustic descriptors, thus describing an acoustic space associated
with the perceptive space.
- The categorization method, in theory free, is an alternative approach to the
previous method and is used when the perceptive continuum hypothesis (and
therefore the homogeneity of the corpus) is no longer valid. This method consists of
grouping them together on the basis of similarity criteria (which can, as the case
may be, be specified: in this case, the categorization is said to be “oriented”), the
different sounds of a corpus, this grouping together occurring with no notion of
order nor numerical value. The most current analysis associated with this method is
a “cluster” type analysis that leads to a hierarchical representation in the form of a
tree ( dendrogram ) describing the structural organization of elements of the corpus
revealed by the experience.
8.5.2. Specificity of the evaluation protocol of auditory HMIs
When the use of sound and implementation of the associated investigation
methods are applied to the domain of HMIs, we then talk of auditory HMIs; the
objective of this application being to facilitate the use, ergonomics and relation
between a user and a system (“system” is taken in the broad sense and includes both
the virtual interfaces and tangible objects or physical spaces in which the user
evolves). In this case, the protocols and experimental factors are influenced by two
important effects that come from the limited context of psychoacoustics, such as
presented in the previous section: context and interaction.
8.5.2.1. Consideration of the context
The degree of consideration of the experimental context defines the
configuration between a so-called laboratory situation ( in vacuo ) and a real situation
( in situ ). This degree can be variable according to the type of application and
experimental means that are implemented to deal with it. For example, concerning
yourself with auditory HMIs in the automobile passenger compartment can lead to
the deployment of a range of contextualizations ranging from a completely
decontextualized situation (passive monophonic listening in a sound booth) to a
realistic situation, which is therefore active (spatial listening on a 3D driving
simulator playing a scenario), even real (vehicle on a track or open road). From an
experimental point of view, the difficulty resides in mastering the different
parameters linked to the context: the more real the test situation, the more complex it
will be and therefore difficult to control. This is the case given that an ideal
experimental framework is defined by mastering all the experimental parameters in
order to identically reproduce it, and, as the case may be, study the possible
Search WWH ::




Custom Search