Voices of Caruso: Cognitive evaluation and acoustic analysis
of reedited editions
Séverine Morange
IJRDA, LAM, Univ. Paris 6/CNRS UMR7190/Ministère Culture et Communication, France
severine75@free.fr
Danièle Dubois
IJRDA, LAM, Univ. Paris 6/CNRS UMR7190/Ministère Culture et Communication, France
ddubois@ccr.jussieu.fr
Jean-Marc Fontaine
IJRDA, LAM, Univ. Paris 6/CNRS UMR7190/Ministère Culture et Communication, France
jmfontai@ccr.jussieu.fr
In: K. Maimets-Volt, R. Parncutt, M. Marin & J. Ross (Eds.)
Proceedings of the third Conference on Interdisciplinary Musicology (CIM07)
Tallinn, Estonia, 15-19 August 2007, http://www-gewi.uni-graz.at/cim07/
Background in cognition. Our exploration of the subjective judgments of Caruso’s voices implies the close
collaboration of both cognitive scientists, mainly psychologists (Dubois, D., 2006) and linguists (Morange, S., 2005)
and physicists and engineers as well (Fontaine, J-M., 2001), within a situated approach of cognition and cognitive
processes, as developed in the LAM. We present here a joint analysis coupling psychological and linguistic evaluations
with acoustic analyses on a unique research object: A Caruso’s piece of song diversely remastered on commercial CDs.
Background in music acoustics. As such, it is motivated by J-M Fontaine’s research on saveguarding and valorising
sound documents of heritage interest. We are especially involved in sound information extraction from analogue
carriers such as records and magnetic tapes, and "restoration" signal processing in order to get at a more precise
definition of what could be “high quality” for sound recordings.
Aims. Up to now, most of the studies have generally been concerned with the development of signal processing
"arising sound quality", but without little feedback from auditors (even as CD consumers). The present pluridisciplinary
study tends to contribute:

get at such an evaluation through the elaboration of a specific experimental protocol;

to determine how a diversity of audience appreciates different versions resulting from different "restoration"
treatment of one single original musical recording.
Method. 32 subjects were selected contrasted on age (related of their different experience of earlier technical
recording devices) and on expertise concerning musical acoustics (acousticians and / or musicians vs. ordinary musiclovers). 11 excerpts of reeditions of an opera record interpreted by Caruso were selected. The listening protocol
involved: Free categorization task, selection of excerpts on preference judgments. Each task involved subjects’ free
commentaries about their choices.
Results. A cluster analysis scaffold by a psycholinguistic processing of the verbal comments of the categories allowed
to identify both commonalities and differences in groupings excerpts by the different groups of the subjects, along a
diversity of criteria, varying according to age and expertise.
Each excerpt can therefore be characterised both according to psychological evaluations and to acoustic analyses
(spectral and temporal analyses).
Conclusions. This study has enabled us to develop the idea that a musical sequence (a lyric voice) is a multi-face
object (cultural, aesthetic, technical, physical), acoustic parameters being diversely linked to the various sensory
experiences and expertises of evaluators.
Implications. Such research and the methodology we developed acknowledge for a better understanding of listening
practices and music-lover assessments here concerned with a specific musical genre (opera), and a diversity of media
technology (analogical or digital records, radio…).
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
E.
Caruso
(1873-1921),
who
was
contemporaneous
with
audio
recording
processing, combined during his artistic life
both stage performance and disc recording.
These
two
types
of
singing
activity
contributed to his fame as well as to the
phonographic industry development.
'restore' the old sound recordings. The high
performance of current DSP have so much
enlarged the possibilities to transform
historical reference sound recordings that it
questions what “pleasant to listen to” means
for present day auditory requirements. From a
simple impulse noise removal to deeper
transformations
by
formant
range
enhancement, adding virtual reverberant
effect, etc, these options question a possible
“betrayal” of the authenticity of the recording
(we don’t deal with this subject here), but also
and more generally the listener’s appreciation
of such modifications.
Caruso's voice was recorded by the best
technical and professional devices at his time,
but nevertheless these recordings could not
benefit from the more sophisticated ones just
available a few years after he died, i.e.
electrical recording developed after 1925.
Therefore, the discs produced were rapidly
considered as limited in quality considering
the rapid evolution of technologies. However,
at this period, it was the first time that
recording processes offer the possibility of
listening
outside
live
lyric
stage
representation. In such conditions, Caruso's
contemporary audience greatly appreciated
the recordings (for instance, the title under
study in the present paper, was reproduced
up to 1 million copies). Successfully, emotion
was conveyed …
The present study contributes to analyse the
relationships
between
studio
operators’
actions on re-mastering and the consumers’
opinions for these new products. In a context
of continual development of restoration tools,
it seems to us there is an urgent need to pay
attention to listeners’ appreciations, and
therefore to benefit from new developments
in cognitive research in order to more
precisely identify auditory evaluations, in
close relations with acoustic analyses and
knowledge.
During one century, the enjoyment of the
public for the singer led to numerous reeditions using the diversity of carriers
available: 78rpm discs, LPs, K7, compact
discs, so to quote only the main physical
vectors of communication. Due to this
recording evolution, – such re-editions
integrate technical devices and human
interventions which more or less transform
the original sound (the sound of original
record).
The pluridisciplinary project we are involved
in integrates a situated cognitive approach of
subjective assessments to the technical and
historical considerations concerned with the
evaluation of Caruso’s voice in contemporary
restitutions. The present study involving
recent insights from cognitive linguistics and
psychology intends to relate technical and
acoustic knowledge to investigations of
present day listening attitudes and quality
judgments, from a diversity of consumers and
listeners.
If auditory requirements evolved through new
technical
developments,
technicians
managing player and audio processing
devices
remain
the
intermediary
who
transcends recording apparatus to go up
towards
the
(their?)
imaginary
initial
performer state.
New proposals have been regularly offered to
play
original
discs
including
diverse
technologies (optical turntable, photographic
capture,…). More recently, electro-acoustic
equipments have been used to improve sound
restitution
by
means
of
equalization
adjustment, filtering device, etc. From the
late-1980s, digital signal processing (DSP)
technologies have been implemented to
Figure 1. Caruso’s sketch of himself. by himself: "Caruso
made the record, the record made Caruso".
2
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
Problematic and aims
quality of such new products derived from
such old recordings.
Such a requirement of connecting the
concrete
constraints
of
technological
developments and the psychological reality of
the audience’s judgments led us to conceive
the present research program integrating
acoustic and psycholinguistic competence.
The involvement of this last domain appears
to be inescapable inasmuch it allows to get at
individual as well as shared representations,
through people’s discourses and comments
about sound quality.
Such an association of different scientific
contributions originated in the following
questions:

how does a
appreciate the
recording?
diversity of audience
quality of a musical

what are the criteria along which people
ground their judgments?

how do they account for such judgments in
their discourses, what words do they use?

what relation can be identified between
their judgments, their verbal comments
and acoustic analyses?
The hypothesis that the relevant cognitive
categories,
through
which
subjective
evaluation is made, is grounded on our
previous work considering that “categorical
knowledge relies on an individual sensory
experience, however shared through the
collective appropriation of linguistic forms and
discourse processing” (Dubois D., 2006).
Therefore, in our common attempt to identify
the relevant properties that define high
quality for a musical excerpt, we set up a
close
cooperation
between
different
disciplinary competences (mainly cognitive
sciences and acoustics) in a joint analysis of
subjective as well as physical description of a
set of recordings.
We
will
mainly
focus
here
on
the
methodological issues, the results being an
illustration of the productivity of such an
approach that could be further applied on a
larger scale to other objects (for more
detailed results, see Morange S. et al., 2005).
In contrast with the main stream in cognitive
sciences, which relies on a priori ontological
categories - processed by an abstract
universal subject -, we are concerned here by
subjects’ evaluations in their diversity of
sensory experience and knowledge, within the
field of what is presently named “situated
cognition” (Dubois D. (dir.), 1991).
Such an approach has already been validated
in other domains in acoustics within the LAM:
Urban
ambiences
and
soundscapes
(Guastavino C., 2003; Raimbault M., 2002;
Dubois D. et al., 2006), warning signals
(Vogel C., 1999), musical instruments
(Castellengo M. et al., 1999), musical timbre
(Castellengo M. et al., 2005), vocal quality in
singing (Garnier M. et al. 2004), or in speech
processing (Morange S., 2005).
Theorical fields
As far as the sonic patrimony is concerned,
and with it all the processes and devices of
conservation, of restoration, as well as of
restitution, it is mainly the know-how and the
skills of the engineers in acoustics which are
mainly involved in remastering (Bassal D.,
2005). The process is therefore under the
control of their (implicit) aesthetic choices,
that may also be influenced by the “look” or
style of record factories or media channels,
and also, more rarely, by some historic
knowledge (Calas M. F., Fontaine J. M. 1996 ;
Fontaine J-M.,
2001),
and
seldom
by
sociologists
(Maisonneuve
S.,
2004).
However,
the
audience
(music-lovers,
amateurs, or media consumers) who should
be first concerned are not questioned. It
seems to us it would be of interest to more
precisely figure out how they evaluate the
Within the present work, the pluridisciplinary
theoretical frame and the methodology it
induces, will be applied to a specific object: A
sample of a diversity of remastered recording
of an extract of an opera piece. Such a
complex object can be described from
different points of view:

3
from a physical point of view, it is at the
same time: A physical event / a sounding
event
/
a
musical
event
-vocal,
instrumental;
CIM07 - Conference on Interdisciplinary Musicology - Proceedings

from a cognitive point of view, it is a multifunctional object: A technical object
through the materiality of the recording / a
sounding object – as acoustic and musical
matter / a cultural object -endowed with
specificities: Musical genre, artistic genre,
historically and geographically marked,
engaging
aesthetical,
cultural
and
meaningful values.
producers,
technicians,
musicians,
musicologists, (…), amateurs are also experts
in
aesthetics,
(…)
aesthetics
being
considered, not as a restricted domain
established as a discipline, but rather as a
practice experienced and developed by any
music lover».
Such criteria were considered in order to test
the influence of subjects’ diversity of previous
sensory experience regarding:
We will therefore investigate this technical
and cultural object in its complexity through a
diversity of approaches.
Methodology
The experimental protocol involves three
steps:

the analysis of such data within cognitive
and linguistic categories to be further

correlated to an acoustic description of the
excerpts.
the type of famous lyric voiceii
Constantin de Chanay H., 2001);

the
type
of
technical
device
involved (78rpm disc vs. digital recording ;
type of player system).
Tasks and instructions. Based on
question sheetiii, the protocol involved:
the elaboration of a listening task
associated to verbal comments (partly
oriented by an open questionnaire);


Experimental setting
As
regularly
processed
within
the
psycholinguistic approach we developed, the
questionnaire itself has been adjusted to the
specificity of the topic and to the subjects
questionedi.
(cf.
a

a free categorization task from the
listening of each excerpt and the grouping
of the all excerpts along similarities and
differences;

followed by the selection of two excerpts
on a preference judgment (the preferred
one vs. the least preferred one);

the subjects’ free comments about their
choices for both the perceptual and
evaluative tasks.
Such categorization task and selection on
preference judgment are common procedures
in psychology, when exploring cognitive
categories. However, the present originality is
given by the connection of such data with a
careful linguistic analysis of the verbal
comments.
Sources.
We
selected
11
recordings
implemented on commercial CDs edited
between 1989 and 2004, of an opera
recording performed in 1907 by E. Caruso
(Pagliacci, R. Leoncavallo -1890). Among
these recordings, one exemplar (excerpt 1) is
considered as an “original replicated” without
any signal processing. For the 10 remaining
remastered
excerpts,
no
processing
information is available.
Analyses
Psycholinguistics methods. We distinguish
two ways of processing subjects’ answers:
Subjects. 32 subjects were selected with a
contrast in age (“younger than 30” and “older
than 60”, related to their possible different
experiences of earlier technical recording
devices), and in expertise in musical acoustics
(acoustician and/or musicians, VS ordinary
music-lovers).
We specially focused on the perceptive
experience, with the idea borrowed from
Maisonneuve, S (2001: 3) that «as well as

a mathematical analysis of the clusters
resulting from the categorization task
defining
similarities
and
differences
between classes and preference judgments
(Poitevineau J., 2004);

a linguistic analysis of the free comments
following
the
categorization
tasks
(selection of the verbal units with similar
semantic references, see Tables 1 and 2)
Acoustic analysis. Our aim was to identify
the treatments that were possibly applied:
Pulse noise attenuation means of digital
4
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
signal
processing,
frequency
response
balancing, dynamic strengthen, reverberation
adding, speed / pitch adjustment, etc.
To define acoustic description characteristics
applied to Caruso’ voice, solo or accompanied
by the orchestra, we carried out selected
sequence analyses of waveform, narrow band
spectrum, averaging 1/3 octave band,
spectrograms ...
These acoustic analyses are to be put
together with the subjects’ categories, their
preferences, and their verbal comments.
background noise, distance effect, lack of
contrast, dated aspect.
Results
Very
detailed
spectrographic
analyses
revealed two procedures used to remove
pulse defects (always present on 78 rpm
records):
Adaptative
signal
processing
(excerpts 8 and 10), and steady filtering
(excerpts 3, 5 and 11).
Acoustic properties
Acoustic analyses have been conducted in a
comparative manner so as to qualify the
different versions.
We established exact identity between two
recordings (excerpts 3 and 5), for which it is
worth noticing that they present a significant
dynamic
reduction
(musical
nuance
attenuation), probably due to secondary
effects induced by signal transformation.
Categorical structures
The mathematical analysis (sorting algorithm,
Barthélemy et al., 1988) allowed the
identification of the grouping of excerpts as
categories:

some categories
subjects;
are
common
to

some are specific to the different groups of
subjects depending on the age and
expertise as shown here below.
Apart from three versions (excerpts 1, 7, and
10), bass frequency (under 120Hz) has been
systematically removed: This very low band
region contains little musical information but
a lot of noise …
all
One-third octave band spectral analyses
reveal filtering and equalization processing
that tend to balance the resonance effect
from
the
original
recording
device
(gramophone horn, sound box, …).
Verbal descriptions
The association of the categories identified
here above with verbal comments revealed
common and diverse features.
This observation can be made at the level of
the cognitive functions that those features
and categories induce. The evaluated object
(the lyric voice) was interpreted by the
subjects as:

an external event (ex: ”bruit de fond au
premier plan”iv - background noise in the
foreground);

a feeling (ex: dryness of interpretationv);

a spatial source (ex: reverberation as in
churchvi);

a temporal source, reminder of a dated
aspect (ex: the needle of an old
phonographvii), or a modern aspect (ex:
CD styleviii).
The addition of reverberation is detected on
two versions (excerpts 8 and 10) and, to a
lesser extent on a third version (excerpt 2).
Such an artificial device endows the singing
voice and the orchestra with sonorousness
and a certain presence.
Among main studied acoustic properties, we
must mention the different playing speed
from one version to the text, with resulting
variance of nearly quatertone (excerpt 4:
Lower pitch, excerpt 7: Higher pitch).
However, the musical sequence did not
permit any subject to detect pitch variation
(singer voice in vibrato regime, orchestral
cluster, filtering effect …).
Finally, we correlated the analyses of the
psychological evaluations (non-verbal free
categorization and verbal comments), with
the acoustic descriptions related to their
possible correspondences. In this way, a lot
This comparison of cognitive and linguistic
data also allows the characterisation of each
excerpt. For example, the same excerpt
(excerpt 1) has been described by most of
the subjects through the criteria of dominant
5
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
properties correspond
filtering, reverberation.
of categories emerged. We present the more
significant ones.
to:
Frequency
area,
cracks,
kHz
Main categories
6
Four categories are specific to:
5

all the subjects;

the excerpts preferred by the most of the
subjects;
4

the younger and the older subjects;
3

the expert subjects.
2
The two first categories will be presented with
tables and figures. The next two will be
mentioned more briefly.
1
1
All subjects classes. A category common to
most subjects was identified through the
grouping of three excerpts (excerpts 3, 5 and
11). Moreover, in their comments, subjects
gave similar references, that we ascribe to
acoustic properties:


2
3
4
5
6
7
8
9
10
11
0
Figure 2. The excerpts Spectrograms present. time
evolution (5 s on X-axis) of frequency components (0 to
6 kHz on Y-axis) which is displayed for each sequence.
Distribution of instrumental harmonic tones and voice
formant areas is revealed means of color or grayscale.
Filtering conditions, formant reinforcement, signal
treatments, pitch definition, are related in detail by such
a useful representation.
the mention of reduced background noise
and of clean aspect, that can be ascribed
to filter addition;
Preferred excerpts classes. Half of the
subjects agreed on a second class through
the grouping of the same two excerpts on the
preferred criterion (excerpts 10 and 11, cf.
Fig 3).
the mention of infra-bass (which can
induce a fullness of the sound or of the
voice), and of balance, can be ascribed to
a low register.
Their commentaries further revealed the
same references as for the first class, which
we attribute, in the same way, to the same
acoustic properties (cf. Table 2). The “clean
aspect” seems to be here a criterion of
positive choice: The excerpt 10 is declicked
and has filter addition, and the excerpt 11
has reverberation added (more on Fig 4).
We present below a summary table of both
the psycholinguistic and acoustic properties
(cf. Tab. 1), and a spectrogram illustrating
the frequency of each excerpt (cf. Fig. 2).
ALL SUBJECTS CLASSES
(excerpts 3-5-11)
Psycholinguistic properties
reduced background noise (reasonable
background noiseix), presence of the bass
(big noise levelx), fullness of the sound
(roundxi), fullness of the voice (sultryxii),
balance (well balancedxiii), clean aspect (it
sounds like CDxiv)
Acoustic properties
infra-bass, non declicking, filter addition,
natural reverberation
Table 1. The Psycholinguistic properties present the
commentaries summed up as semantic references (and
not as verbatim) describing a notion (ex: background
noise),
qualified
and
evaluated
(ex:
reduced,
important…). A verbatim is given as an example of one
explicit subject’s comment (here in italics). The Acoustic
Figure 3. The histogram presentation of all subjects
answers to the question concerned with the selection of
the preferred excerpt: X-axis corresponds to each
excerpt; Y-axis corresponds to the number of subjects
6
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
having chose such excerpts. It is interesting to point out
2 extreme excerpts: The non appreciated one without
any treatment (noisy but complete information present:
Excerpt 1), and the preferred version (very “clean” but
partial information is lost: Excerpt 11).
background noise as a cue pointing to the
historical characteristic (ex: very old recordxx)
of the recording in association with the
perception of the distance of the voice and
the orchestra. These two groups can be
explained in relation to identical acoustic
properties, but with different meanings given
to it by the different groups of subjects, with
high and low frequency ranges for younger
and older subjects respectively.
PREFERRED EXCERPTS CLASSES
(excerpts 10-11)
Psycholinguistic properties
reduced background noise (with the less
interferencesxv), presence of the bass
(enough bassxvi), balance (voice and music
at the same levelxvii), fullness of the sound
or the voice (rounder resonancexviii)
Acoustic properties
lot of infra-bass, non (11) declicking (10),
non (10) filter addition (11), natural
reverberation (11), added reverberation
(10)
Experts and non experts subjects’
classes. Experts’ grouping on their side
(excerpts 8 and 10) relies on the
identification of technical devices such as
background noise and reverberation (ex:
reverberating room, low cracklesxxi), probably
attributed
to
declicking
and
added
reverberation processing. Whereas experts
tend to pretend to a more “objective”
evaluation of the sound per se, in contrast,
non-experts subjects were mainly grouping
through an evaluation of their feelings or
through the reminder induced by the
stimulation.
Table 2. Cf. Table 1 for the legend.
Conclusion
The preservation and restoration of musical
recordings involve a necessary collaboration
of physical sciences and signal processing
knowledge and human sciences. The first
ones are obviously concerned with the
recordings as a physical object whereas the
latter deals with the evaluation of them as
cultural objects that may vary across time
and across the diversity of individuals.
Then, the diversified properties identified at
the issue of the cognitive and linguistic
analyses
revealed
the
“heterogeneous
modalities of our inclinations“ (Hennion A.,
2002). In other words, the different listening
practices depend on our own attitudes and
motivations (technical, aesthetic, hedonistic),
and on our previous knowledge and
experience as ordinary listener or expert, that
may give different meanings to the “same”
sound.
Figure 4. Frequency analysis 1/3 octave band. To bring
acoustic measurement closer to perceptive consideration,
sound level (extraction and average energy signal done
during sequences duration) is displayed (in dB) in a
range of frequencies. On a logarithm scale, audible
spectrum (20 Hz to 20kHz) is presented means of 3
equal bandwidth in each octave (doubling of frequencies
region partition). We observe filtering effect of low and
high frequencies (excerpt 10 and 11 vs. excerpt 1) and
enhancement effect on medium frequency range (excerpt
10 and 11 between 160 and 500 Hz).
Age’s subjects classes. Most subjects
agreed on excerpt 1 through the mention of
background
noise,
but
they
focused
differently depending on their age.
Younger listeners were grouping excerpt 1
with excerpt 9, with background noise as a
technical
impairment
(ex:
hiss
interferencexix),
whereas
older
subjects
grouped excerpt 1 with excerpt 6, on
Lyrical voice thus appears as multi-face
object, according to a diversity of sensitive
experiences
and
expertises
of
the
participants. This object can be evaluated as:
7
CIM07 - Conference on Interdisciplinary Musicology - Proceedings

a technical object (that may be analysed
by physical parameters, and that can be
changed);

a sounding object (that may be analysed
through acoustic properties);

a cultural object (listened and evaluated as
revolving to a modern or dated document).
From the acoustic point of view, “the
recording of the voice is problematic for
technicians in spite of a more and more
sophisticated material that is not always
efficient: How to make restitution of an opera
piece? Do we have to find out a new way of
listening for musics which were created when
the diffusion was only in direct? (…). Indeed,
productions suffer from the historical lag of
interpreting!” (Bertier M. 1986: 227).
Bertier,
M.
(1986).
Les
conditions
de
l’enregistrement et de la diffusion de la
voix chantée, in La voix dans tous ses
éclats, Actes du colloque, 27-29 avril 1985,
Centre G. Pompidou, Année Européenne
de la Musique, 224-228.
Calas,
M. F., Fontaine, J. M. (1996). La
conservation des documents sonores,
CNRS ed., Paris.
Castellengo, M., Besnainou, C., Dubois, D. (1999).
Acoustic quality of musical instruments
and categorization, in Actes du “137th
Meeting of the acoustical society of
America and
forum acusticum”, Berlin
14-19 mars 1999, 12-17.
Castellengo, M., Dubois, D. (2005). Timbres ou
timbres?
Propriété
du
signal,
de
l’instrument,
ou
construction(s)
cognitive(s)?, (CIMO5), Cahiers de la
Société Québécoise de Recherche en
Musique, to be published.
These observations converge with our own
questions: Which treatment to choose for a
“good” restoration of a musical recording,
from which support (cylinder, disc), on which
storage tool (analogical, digital), and last, but
not least question, for which audience (musiclovers, young / old, radio listener….)?
Constantin de Chanay, H. (2001). La voix
d’opéra: Sémiologie et rhétorique, in
Puissances de la voix. Corps sentant,
corde sensible, Badir S., Parret H. (dir.),
Limoges, PULIM, 91-110.
Dubois, D. (dir.) (1991). Sémantique et cognition.
Catégories, prototypes, typicalité, CNRS
ed. Sciences du Langage.
In conclusion, we thing that it worth
improving our knowledge of the listeners’
practices and tastes, and its relations to their
acquaintance with the artistic style (opera),
and with the physical object (disc, CD, radio).
Therefore, the procedure we suggested here
can be further developed and generalised in
order to more largely explore the recording
"authenticity", that professionals involved in
sound heritage may wish to promote.
Dubois,
D. (dir.) (1997). Catégorisation et
cognition: De la perception au discours,
ed. Kimé.
Dubois,
D. (2006). Sens commun et sens
commun:
Expérience
sensible,
connaissance(s)
ou
doxa ?,
Langue
française, to be published.
Dubois, D., Guastavino, C., Raimbault, M. (2006).
A cognitive approach to soundscapes:
Using verbal data to access auditory
categories, Acta Acustica & Acustica, 92, 6,
865-874, submitted.
We expect to have provided a new productive
approach to the analysis of audio perception
in general, and to musical listening in
particular.
Fontaine, J-M. (2001). De
documents sonores.
Accès, dégradation,
(Conservation
et
patrimoine culturel),
46.
Acknowledgments.
The
authors
wish
specially to thanks C. Vogel for his help with
acoustic analyses, and J. Poitevineau for the
mathematical analyses.
la restauration des
Les disques noirs:
restauration, CORE
restauration
du
n°10, juin 2001, 36-
Garnier, M., Dubois, D., Poitevineau, J., Henrich,
N., Castellengo, M. (2004), Perception et
description verbale de la qualité vocale
dans le chant lyrique: Une approche
cognitive, JEP2004, Fez-Maroc.
References
Bassal, D. (2005). La pratique du Mastering, Mac
Music, online, http://www.macmusic.org
Guastavino, C., (2003). Etude sémantique et
acoustique de la perception des basses
fréquences dans l’environnement sonore
Barthelemy, J. P., Guénoche, A. (1988). Les arbres
et les représentations des proximités,
Masson, Paris.
8
CIM07 - Conference on Interdisciplinary Musicology - Proceedings
urbain, doct. Acoust., dir. Dubois D., Paris
6.
For further information on the topic cf. Morange S.,
Candea M., 2007.
i
Hennion, A. (2002). Ce que ne disent pas les
chiffres … Vers une pragmatique du goût,
Colloque Les publics, 28-30 novembre
2002, DEP-Ministère de la culture / OFCEFondation
nationale
des
sciences
politiques, 7 p.
Would the performance be the reflect of a historical
time? While vocal and instrumental techniques greatly
changed?
ii
1) Please, successively listen to the 11 extracts. What
are your first impressions with regards to the nature of
the recording of the different excerpts?
iii
Jones, M.R., Yee, W. (1994). L’attention aux
événements
auditifs:
Le
rôle
de
l’organisation temporelle, in Penser les
sons. Psychologie cognitive de l’audition,
Mc Adams S., Bigand E. (dir.), Psychologie
et sciences de la pensée, 75-121.
2) Now, group together the excerpts that sound similar
and group together in different clusters those you find
different. You may create as many clusters as you wish,
with different numbers of extracts in the clusters.
3) Please, assign the results of your selection and try to
comment it (table).
Maisonneuve, S. (2001). De la « machine parlante
à l’auditeur, Terrain, n°37, Musique et
émotion, 18 p.
4) Now, select 2 extracts among the 11: the one you like
most, the one you like least. And try to comment it.
We deliberately give the French verbatim inasmuch the
English translation is seldom straightforward. For
example, for the French word “chuintement”, a rapid
scan of several English dictionaries gives “swish”,
“(gentle or spatter) hiss”, “screech”…
Morange, S. (2005). Approches structurale,
prosodique, psycho-cognitive de quelques
propriétés linguistiques subjectives de la
parole: L’exemple de trois Montmartrois
âgés, doct. Linguist., dir. M. A. Morel, Paris
3.
iv
v
“Sécheresse de l’interprétation”.
Morange, S., Candea, M. (2007). Aux frontières de
l’écoute. Réflexion sur la construction des
variables pertinentes dans la mise en
place
des
tests
de
perception,
in Frontières, dir. Morel M. A. et Delomier
D., Bibl. de l’Information Grammaticale,
Peters, to be published, 10 p.
vi
“Reverberation comme à l’église”.
vii
“L’éguille d’un vieux phonographe”.
Morange, S., Fontaine, J-M, Vogel, C., Poitevineau,
J., Dubois, D. (2005). Appartenances
catégorielles d’un objet complexe, la voix:
Approches
linguistique
et
psychocognitive, 3ème Journée du Sensolier, 7
oct., CNRS - CRP, Ivry.
Poitevineau, J. (2004).
statistiques
par
psychologie: Aspect
prescriptif, Math. &
(3), 5-25.
“Genre CD”.
viii
L’usage des tests
les
chercheurs
en
normatif, descriptif et
Sci.hum n°167, 2004
Raimbault, M. (2002). Simulation des ambiances
sonores urbaines: Intégration des aspects
qualitatifs, doct. Méca. therm. génie civil,
Univ. d’archit. Nantes, dir. Péneau J. P.
Vogel, C. (1999). Etude sémiotique et acoustique
de l'identification des signaux sonores
d'avertissement
en
contexte
urbain,
doct. Acoust., dir. J-D Polack, Paris 6.
9
ix
“Bruit de fond acceptable”.
x
“Gros volume”.
xi
“Rond”.
xii
“Chaude”.
xiii
“Bonne balance”.
xiv
“Son de CD”.
xv
“Le moins parasité”.
xvi
“Assez grave”.
xvii
“Voix et musique au même niveau”.
xviii
“Sonorité plus ronde”.
xix
“Chuintement parasite”.
xx
“Très vieux vinyle”.
xxi
“Salle réverbérante”, ”peu de craquements”.