Information Technology Reference
In-Depth Information
ows
coming from all sensors, except the signals coming from the EEG electrodes. These
variation speeds are translated into energy by a direct control of the volume,
spectral richness, or pitch of a sound. As regards the heart and breath sensors, the
frequency of the cycle is evaluated (heart rate or inhalation-exhalation period), and
the variations of this frequency are also translated into volume, spectral richness, or
pitch variations. Concerning the signals from the EEG electrodes, the evolution of
energy ratios between wave bands re
In speci
c terms, we compute in real time the variation speeds of the data
ecting typically vigilance states is soni
ed in
the same way.
11.2.2 Interaction Between Vigilance and Musical Synthesis
In addition to the production of sound directly from physiological signals, the
energetics approach also involves a higher level of abstraction by analyzing the
listener
flow, using harmonic, rhythmic, and
spectral rules. The idea here is to control the dynamism of this musical flow
according to the listener
'
s arousal in order to generate a musical
s arousal.
All the physiological data are therefore integrated into one continuous parameter
representing the listener
'
is arousal, and this parameter is then redistributed over the
whole set of musical generation parameters. It is a
'
M
!
1
!
N
mapping.
A study carried out by Pierrick Legrand and Fr
é
d
é
rique Fa
ï
ta Ainseba (Vezard
et al.
2011
) shows the link between speci
c musical composition parameters, the
energy evoked by the resulting generated music, and the arousal level induced by its
listening.
These parameters, as they are used in the MuZICO environment, have either
discrete or continuous values and are either de
ned by adjectives (classes) or by
numerical values. For example, the spectrum of a sound can take values from muted
to bright, which corresponds to speci
c energy values that are measurable in the
signal and are linked to its spectral centroid. The other parameters related to energy
in music are as follows: the tempo of the music, the attack of the notes
dynamic
envelope, the pitch of the notes, the number of instruments playing simultaneously,
the more or less
'
sound of the instruments, the percussive or sustained
aspect of sounds, as well as the number of simultaneous notes (Livingstone et al.
2010
).
All these parameters are characterized by bounded intervals used by MuZICO to
generate music.
Let be
“
natural
”
the bounds of the intervals of values taken by musical
parameters, corresponding to low and high energies evoked by the music, respec-
tively (see Table
11.1
). The parameters that have undefined units are translated by
MuZICO into various units according to the algorithms used to produce the sounds
(synthesis, sample playing, audio-effect control). For example, one can translate the
brightness of a sound by a particular setting of the modulation index in a FM
V
E
min
and
V
E
max
