PDF Loudspeaker sound quality: comparison of assessment -[PDF] Subjective Assessment of the Spatial Attributes of Reproduced Sound







Chose PDF link
PDF :1 PDF :2 PDF :3 PDF :4 PDF :5 PDF :6 PDF :7 PDF :8



source https://pdfs.semanticscholar.org/49b0/f0dbb6acb853af8106d54aba7f1b55eb6298.pdf

Loudspeaker sound quality comparison of assessment procedures V Koehl and M Paquier LISyC EA 3883, 6 avenue Victor Le Gorgeu, CS 93837, 29238  A paired comparison is an easy and reliable way to assess loudspeaker sound quality Subjects could directly listen to sound radiations by loudspeakers Subjective assessment,









LOUDUN - France

[PDF] Téléchargez la plaquette - Loudun

Wonder Horror Loudun in France is a small town near Poitiers built on a hill surrounded by wide plains “A cluster of belltowers surrounded by fields” It was built  sacLt^dLw ~ *2U—ii L'éxecution de Urbain Grandier à Loudun au couvent des Ursulines de Loudun dans

LOUDUN - France [PDF] Téléchargez la plaquette - Loudun

[PDF] Wonder-Horror Loudun in France is a small town near - UGALcom

Wonder Horror Loudun in France is a small town near Poitiers built on a hill surrounded by wide plains “A cluster of belltowers surrounded by fields” It was built 
PDF

L'envoûtement de Loudun - Érudit

sacLt^dLw ~ *2U—ii L'éxecution de Urbain Grandier à Loudun au couvent des Ursulines de Loudun dans les années maître de la France Ses querelles 
PDF

[PDF] Bassin d'emploi de Loudun - Cap métiers

En 2014, le bassin d'emploi de Loudun compte 4 685 salariés sur 7 518 actifs occupés Sur la période ECP France, Commune de Loudun, Communauté  
PDF

[PDF] CARTE GÉOLOGIQUE DE LA FRANCE A 1/50000 - BRGM

LOUDUN est recouverte par les coupures suivantes de la carte géologique de la France à 1 80 000 à l'ouest SAUMUR (N° 119) à l'est LOCHES (N° 120)
PDF

[PDF] AIP for FRANCE - SIA

LOUDUN AD 2 LFDL ATT 01 ATTERRISSAGE A VUE Visual landing Ouvert à la CAP Public air traffic LFDL VAR 0° (15) Aides lumineuses NIL Lighting 
PDF

[PDF] Téléchargez la plaquette - Loudun

Thouars LOUDUN C'EST DANS LA CITÉ PROTESTANTE DE LOUDUN QUE NAÎT THÉOPHRASTE RENAUDOT France en 1631 En 1925, au restaurant
PDF


Loue à Saint Gervais les Bains (haute Savoie)

[PDF] Le logement des saisonniers en Haute-Savoie

Loue à Saint Gervais les Bains (haute Savoie) Au pays du Mont Blanc, Bel appartement type F2 pour 4 personnes dans un beau chalet au rez de jardin, hotel saint gervais les bains hotel a saint gervais haute savoie hotel saint gervais mont blanc hotel 3 etoiles saint gervais les

Loue à Saint Gervais les Bains (haute Savoie) [PDF] Le logement des saisonniers en Haute-Savoie

[PDF] Loue à Saint Gervais les Bains (haute Savoie) - Cairns

Loue à Saint Gervais les Bains (haute Savoie) Au pays du Mont Blanc, Bel appartement type F2 pour 4 personnes dans un beau chalet au rez de jardin,
PDF

[PDF] hotel la garissade - Hotel La Fleche d'Or

hotel saint gervais les bains hotel a saint gervais haute savoie hotel saint gervais mont blanc hotel 3 etoiles saint gervais les bains hotel de charme saint gervais 
PDF

[PDF] glisse glisse - Saint-Gervais

Le Fayet I saint gervais I Le bettex I saint nicoLas de véroce I Le prarion Fr I gb contemporary Savoie chalet is ideally located just 250 m from the ski lifts and the Ce refuge est situé aux portes de la haute montagne, aux abords du glacier de Bion Agence de locations de vacances appartements et chalets à louer
PDF

[PDF] Le pont de Saint-Gervais

service RTM de Haute Savoie, Mairie et OT de Saint Gervais, WWF Megève, Passy, Praz sur Arly, Saint Gervais les Bains et Sallanches, la Communauté de Communes Pays du souhaitent louer leur meublé via le service tourisme
PDF

[PDF] GUIDE pratique - Forfaits de ski - Ski Saint-Gervais

Gare TGV ITALIE SUISSE Haute Savoie A40 A410 A41 A40 A43 bienvenue àsaint gervais Mont Blanc WELCOME TO SAINT GERVAIS MONT BLANC 
PDF

[PDF] Les Thermes de Saint Gervais les Bains La Balnéothérapie de

74170 Saint Gervais les Bains Tél 04 50 47 76 08 Site guide piscine haute savoie piscine de saint gervais les bains 2175 P Centre Sportif du 
PDF

[PDF] Le logement des saisonniers en Haute-Savoie

Cette étude est l'occasion, en Haute Savoie, de mieux cerner les modalités de logement des dont 3 stations thermales Saint Gervais les Bains, Évian les Bains, réticence des propriétaires à louer à des saisonniers (crainte de 
PDF


Loué Christophe

[PDF] Compte-rendu de réunion - Gest'eau

Biographie Titulaire d'un Doctorat en Sciences de l'Education de l'Université de Nancy 2, Christophe Loué est professeur chercheur depuis 2005 et Délégué à  Biography Christophe Loué is Associate Professor at Novancia Business School Paris, a business school specializing in entrepreneurship and commerce run  l'entrepreneuriat Christophe Loué focalise ses recherches sur

Loué Christophe [PDF] Compte-rendu de réunion - Gest'eau

[PDF] Christophe Loué - Novancia

Biographie Titulaire d'un Doctorat en Sciences de l'Education de l'Université de Nancy 2, Christophe Loué est professeur chercheur depuis 2005 et Délégué à 
PDF

[PDF] Loué Christophe - English version - Novancia

Biography Christophe Loué is Associate Professor at Novancia Business School Paris, a business school specializing in entrepreneurship and commerce run 
PDF

Les compétences de l'entrepreneur marocain : validation - Érudit

l'entrepreneuriat Christophe Loué focalise ses recherches sur l'entrepreneuriat, et notamment l'identification des compétences entrepreneuriales, ainsi que la 
PDF

[PDF] Les compétences de l'entrepreneur marocain - E-polytechnique

Christophe LOUÉ Titulaire d'un Doctorat en sciences de l'éducation de l' Université de Nancy 2, Christophe Loué est professeur chercheur depuis 2005 et  
PDF

[PDF] 1 LES COMPETENCES - Je crée dans ma région

1 LES COMPETENCES ENTREPRENEURIALES DEFINITION ET CONSTRUCTION D'UN REFERENTIEL Eric Michaël LAVIOLETTE et Christophe LOUE
PDF

[PDF] Massiera Loue - Académie de l'Entrepreneuriat

approche par les facteurs motivationnels Philippe MASSIÉRA, PhD EDC Paris Business School Christophe LOUÉ, PhD NOVANCIA Business School Paris
PDF

[PDF] Référentiel de compétences entrepreneuriales - Social Business

tout en s'appuyant sur ceux développée par Laviolette, Loué, Baronet, Bayad, LAVIOLETTE Eric Michaël et LOUÉ Christophe (2006), « Les compétences 
PDF

[PDF] Christophe TINSEAU - tinseaucom

16 Le VEHICULE est loué avec le plein de carburant Les tarifs n'incluant pas le carburant, si le VEHICULE n'est pas restitué avec le plein, le locataire devra 
PDF

[PDF] Compte-rendu de réunion - Gest'eau

Commission Locale de l'Eau du SAGE Haut Doubs Haute Loue Réunion du propriétaires riverains de la Loue et du Lison X M Christophe SIRUGUE
PDF


Loué Christophe - English version

[PDF] housing guide - OECD

Biography Christophe Loué is Associate Professor at Novancia Business School Paris, a business school specializing in entrepreneurship and commerce run  PDF The First Epistle of H N a Crying Voyce of the Holye Spirit of Loue Translated Out of Base Almayne Into English The new era Chihpen woman required

Loué Christophe - English version [PDF] housing guide - OECD

[PDF] Loué Christophe - English version - Novancia

Biography Christophe Loué is Associate Professor at Novancia Business School Paris, a business school specializing in entrepreneurship and commerce run 
PDF

[PDF] Read Book \\ The Done Thing (Paperback) &# PZQRYZ5RZKO0

[PDF] The First Epistle of H N a Crying Voyce of the Holye Spirit of Loue Translated Out of Base Almayne Into English The new era Chihpen woman required reading books Chihpen woman Liu Jieli financial surgery(Chinese Edition)
PDF

[PDF] Recommendations English version - Euromed Heritage

Présentations de Abdelaouahab ZEKAGH et Christophe BOULEAU déduire des revenus de location provenant de biens résidentiels loués pour un 
PDF

[PDF] Entretien La Ribot avec Christophe Khim (english translation)

Entretien de La Ribot avec Christophe KIHM – le 16 mars 2010 Christophe Kihm Quelle est Sebastián, qui est la personne qui loue ces chaises, me connaît
PDF

[PDF] The Goat Woman of Largo Bay - New America

[PDF] The First Epistle of H N a Crying Voyce of the Holye Spirit of Loue Translated Out of Base Almayne Into English the collection stacks of children's literature Chunhyang Qiuyun 12 Children's Literature 2004(Chinese Edition)
PDF

[PDF] Download eBook &# General Directions for Fulfilling - New America

251 x 175 mm Language English The First Epistle of H N a Crying Voyce of the Holye Spirit of Loue Translated Out of Base Almayne Into English (1574)
PDF

[PDF] Moving to and Living in France - CA Britline

Jan 3, 2018 · English Language Media cannot accept responsibility for claims of advertisers or According to Chris Sears, at the station in Bergerac they 
PDF

[PDF] housing guide - OECD

06 12 64 79 31) or Christophe Poinot A Louer( avendrealouer Translation of standard contract for rental of unfurnished dwellings (in English )
PDF


[PDF] Loudspeaker sound quality: comparison of assessment procedures

Loudspeaker sound quality comparison of assessment procedures V Koehl and M Paquier LISyC EA 3883, 6 avenue Victor Le Gorgeu, CS 93837, 29238 
PDF

[PDF] Loudspeaker sound quality: comparison of assessment procedures

A paired comparison is an easy and reliable way to assess loudspeaker sound quality Subjects could directly listen to sound radiations by loudspeakers
PDF

[PDF] General methods for the subjective assessment of sound quality - ITU

Subjective assessment, audio quality, stereophonic image quality, front image interpretation of the results, so that the best possible comparisons may be Tests which do not include the loudspeakers (or headphones) as part of the system
PDF

[PDF] Subjective assessment of audio quality - EBU Tech

EBU Technical Review Winter 1997 Hoeg et al jective assessment of sound programme quality listening in a certain room using loudspeaker pre
PDF

[PDF] Assessment of audio quality and experience using - ICA 2016

Using a full paired comparison test paradigm, preference ratings were Sound quality evaluation for loudspeaker based systems may follow a “spatial and 
PDF

[PDF] Views on Sound Quality - ICA 2016

The talk presents an overview of research on Sound Quality evaluation The talk reviews and discusses different research streams regarding binaural re synthesis (“binaural audio”) or loudspeaker based sound field synthesis, there is no
PDF

[PDF] Perceptual Audio Evaluation - Delta

Audio quality assessment Direct audio Why is the perceived sound quality of a loudspeaker in two different positions in the same room are compared in a
PDF

[PDF] Subjective Assessment of the Spatial Attributes of Reproduced Sound

Sound systems delivering enhanced Spatial reproduction compared with conventional two channel stereo Sound quality assessment methods of the past 30 issues when assessingthe qualitiesof loudspeakersor low bit ratecodecs
PDF

Loudspeaker sound quality: comparison of assessment

  • Acoustics 08 Paris

Loudspeaker sound quality: comparison of assessment

  • procedures

Koehl and M

Paquier

LISyC EA 3883,

  • 6 avenue Victor Le Gorgeu,

CS 93837,

  • 29238 Brest Cedex 3,
  • [email protected]

Acoustics 08 Paris

In listening tests involving different loudspeakers and aimed at assessing the sound quality of these soundreproducing systems,

the level is generally adjusted to compensate for differences in sensitivity

The loudness

sensation must be alike for each system under test

Because of the non-stationary nature of the musical signals

used as test material in loudspeaker ratings,

loudness assessment by using the current models (Zwicker,

  • ) remains slightly inaccurate

In practice,

loudness is often equalized by ear by the experimenter

study deals with the comparison of various test procedures

The first experiment was a paired comparison of

loudspeakers where short-duration stimuli were presented to listeners for preference ratings

In the second

  • experiment,

the same listeners were allowed to switch,

  • at any time,

from one loudspeaker to another one so that

the proposed stimuli were longer

In both experiments the loudness was equalized by the experimenter

However,

under normal listening conditions,

the listener is usually free to adjust by himself the reproduction

At last,

  • in a third experiment,

the listeners had the opportunity,

  • at any time,
  • to not only switch from one
  • system to another one,

but also adjust the loudness of the stimuli

Introduction

  • 1 Reproduction level

A subjective evaluation of loudspeakers is a difficult and

  • time-consuming task

The parameters to be controlled are

numerous and the results are often very context-dependent

Though the Audio Engineering Society and the

International Electrotechnical Commission have both

provided recommendations for loudspeakers listening tests

till now no standardized technique has been adopted

Nevertheless,

  • in sound quality evaluation,
  • some existing

comparison procedures appear as the most achievable and

  • reliable

Clear guidelines for these tests are as follows: i)

the loudspeakers under test are usually presented as pairs,

which facilitates the comparison

  • ii) various sound events
  • have to be tested

iii) loudness must be equalized over all of

the loudspeakers involved in a session

  • iv) short stimuli

must be preferred and v) the loudspeaker positions must be

exchanged throughout the experiment to avoid positional

  • effects

Because of the influence of loudness on sound quality

  • judgments,

a common prerequisite is to check that the

perceived reproduction level is alike for all of the

  • loudspeakers under test

The current loudness models

(Zwicker [5] or Moore [6]) have been acknowledged for

  • stationary sounds,

but not for the musical signals commonly

used for loudspeaker comparison

Therefore,

loudspeakers listening tests the loudness is generally

matched by ear by the experimenter himself or by some

  • expert listeners

This matching is one of the biggest differences between the

listening tests carried out in laboratory and real-life

  • comparisons

Therefore,

it appeared interesting to design a

test where the listener was free to set the reproduction level

to his convenience and to compare the results to loudnessmatched experiments

All of these constraints make listening tests very far from a

  • realistic listening situation

Under realistic conditions of

  • comparison,
  • a listener (audiophile,
  • sound engineer…)

usually listens to various types of music (generally rather

long excerpts) and can even vary the reproduction level

Such an approach is totally different from the listening tests

carried out within a laboratory

  • 2 Presentation method

In paired comparisons,

  • the subject listens to either

successive stimuli or alternate ones:

This study was aimed at comparing three different

  • assessment procedures,

all based on paired comparisons

Four different loudspeakers were under test to determine

whether their preference ratings are affected by the

  • assessment procedure

The first method is the so-called A-B comparison,

where the excerpt to be listened to is at first

presented over the loudspeaker system,

  • then over the other one,

They have to be

compared one just after the other,

  • because of the

short human auditory memory [7]

This procedure

appears as the most exact way to compare the

  • same excerpt
  • as recommended in [8],
  • should not exceed about 5 s

It ensues that this

type of listening is far from natural conditions for

  • loudspeaker comparison

The second method is an alternate listening and

enables the presentation of longer musical

  • excerpts

The comparison is made by switching

from one system to the other one

  • this is done by

either an operator [3] or the listener himself [7]

This also enables one to test more than two

loudspeakers in a single trial [9]

This approach is

closer to the way music is listened to in real life

Paired comparison

Paired-comparison appears as an easy and reliable way to

estimate loudspeaker sound quality

According to Toole [3],

comparisons must be as quick as possible to ensure

maximum discrimination and minimum variability in the

  • judgments

Even when absolute judgments are desired,

loudspeakers are presented by pairs for maximum

  • discrimination

This procedure is referred by IEC as paired

  • ratings [2]

In listening tests involving paired comparison,

the two stimuli are generally matched in loudness

A strong

relationship between the perceived sound quality and the

reproduction level was established in [4]

Acoustics 08 Paris

N ( N − 1)

  • 3 Design of experiments

Taking these considerations into account,

three pairedcomparison experiments were designed

Successive and

  • alternate presentations,
  • described in 2
  • were proposed to
  • the listeners

Even though the second presentation method

appears as more natural for an average listener,

  • the fact that

he cannot control the reproduction level could be

  • disturbing

In order to compare realistic comparison

approach to laboratory listening tests,

  • a third assessment
  • procedure was designed

For that purpose,

  • the listener was

free to adjust the reproduction level before choosing his

  • preferred loudspeaker

The result of a comparison of two loudspeakers is strongly

dependent upon the content of the excerpts used as test

  • material

The pieces of music chosen for loudspeaker

  • comparison were:

Leonard Bernstein,

West Side Story"

  • symphonic
  • orchestra

Ben Harper,

I want to be ready"

  • human voice
  • and acoustic guitar

George Gershwin,

Rhapsody in blue"

  • piano solo

They were extracted from CDs as 16-bit,

  • 1-kHz wave
  • format files

They were selected upon their ability to reveal

spectral and preferential differences between different

  • loudspeakers

The excerpts chosen for session involving an

alternate presentation lasted 30 s

They were shortened to

  • 5 s'for consecutive presentation

Experiments

Three different assessment procedures were designed to

gradually get close to real-life comparison approach

all involved the same loudspeakers and music excerpts

Loudspeakers were presented in monophonic reproduction

because it has proven to give more discriminating quality

ratings than stereo or multichannel restitution [10]

listeners had to take part to three sessions where the

loudspeakers were presented by pairs

The same excerpt

was played over the loudspeakers under test

The subject

was asked to indicate his preferred loudspeaker

A session consisted of the 18 trials required to assess the 3

excerpts on the 4 loudspeakers: the first excerpt was

proposed for the first 6 comparisons,

  • the second excerpt

was heard for the next 6 comparisons,

  • and so on

It is worth

noting that the three excerpts were presented in a random

  • order over a session
  • 3 Test procedures
  • 1 Measurement scale

A test was divided into three different sessions described

A preference rating scale had to be chosen to assess the

  • answers

According to Toole [11],

  • such a scale has to be

used when the comparison of sounds relies on a relative

basis rather than an absolute one

This kind of scale is

considered by Jason [12] as intermediate in difficulty

between a raw statement of preference and the IEC scale

based on differences between fidelity judgements [2]

The continuous scale was divided into four equally wide

intervals delimited with the labels indicated in Fig

  • (translated from French)

Session 1 was assumed to be the most repeatable

  • and reliable
  • the 5-s stimuli were used

In one trial,

the excerpt was successively presented on the

  • loudspeakers,

A and B

The subject was allowed to

listen to the pair of stimuli as many times as

needed before expressing his opinion through the

  • measurement scale

Loudness was matched for this

  • session

Session 2 was meant to be more consistent with

  • everyday life listening

Longer stimuli (about 30 s)

were proposed to the listener,

  • who had the
  • opportunity to switch,
  • at any time,

loudspeaker A to loudspeaker B

He was allowed

to listen to the excerpt and switch between both

loudspeakers as many times as needed to make his

  • opinion

The excerpt could be played from its

beginning or from any other point chosen by the

  • listener in the timeline,
  • and the listening could be
  • interrupted at any time

The question asked to

listeners was always the same and was about their

  • individual preference
  • loudness was also matched
  • for this session

Session 3 was exactly the same as session 2,

from the fact that loudness was not matched

listener was allowed to vary the reproduction level

of each loudspeaker under test by using a

  • dedicated fader

The fader half-lift corresponded to

  • the matched loudness,
  • and the level could be
  • varied from
  • - 6 dB to + 6 dB around this value
  • the beginning of each trial,
  • the fader value was

randomly set within these limits

The listener was

  • 1 Answering scale of the assessment procedures

The listener had to answer by moving a sliding cursor along

  • the preference axis

The same scale was used during the

  • whole test
  • 2 Stimuli

A stimulus denotes an excerpt reproduced over a

  • loudspeaker

Four loudspeakers (namely A,

C and D'for

the rest of the study) from different makers and assumed to

be of about the same quality were then compared by pairs

  • during the test

A trial begins with the presentation of two stimuli and ends

  • with their ratings

if a single excerpt gives N = 4

  • stimuli,

the number of trials for this excerpt is:

Acoustics 08 Paris

told to set the reproduction level at a comfortable

value before making his comparison

All of these 3 sessions were randomly presented to the

  • listener over a test,
  • which took about 1 h

Each session was

preceded by a 5-min pre-test to familiarize the listener with

  • the answering interface

The instructions were given orally and in written form

Listeners were explicitly told to assess the stimuli

according to their preference independently of their taste

  • for the musical content
  • 4 Loudness matching

In this test,

loudness was matched for sessions 1 and 2: a

continuous pink noise was first used to roughly match the

  • loudspeakers

The level was objectively equalized through

adjustment of the gain control of each loudspeaker till

getting 80 dB(B) at the listening position

adjusted subjectively by three expert listeners so that the

loudness sensation was the same for the 4 stimuli issued

  • from a given excerpt

Each music excerpt was reproduced

so that the level was close to the listening level preferred by

  • the average listener [2]
  • 2 Listening room arrangement for monophonic
  • loudspeaker comparison

In agreement with IEC recommendations [2],

  • loudspeaker

positions were exchanged throughout the experiment to

compensate for the positional influence over the preference

  • ratings

The 4 loudspeakers were swapped between two

listeners so as to test all of the 24 possible combinations of

  • positions
  • 5 Loudspeaker locations

Since the purpose of this study was to get close to real-life

conditions for loudspeaker comparison,

  • the listeners could

directly listen to sound radiations by loudspeakers rather

than to recordings of these loudspeakers

Using direct

  • presentation,

two loudspeakers cannot be set exactly at the

  • same position for comparison

The effects of loudspeaker

positions can be higher than the subjective differences

between the loudspeakers themselves [7]

On the other

Bech [13] noticed that,

  • for most loudspeakers,

timbral quality of reproduced sounds is usually unaffected

by changes in position within a radius of approximately

  • 6 Listeners

The listeners involved in the test consisted of 5 women and

  • 43 men aged from 20 to 60 years
  • the average age was 28

Most of them could be considered as trained listeners since

they were sound engineering students or professionals

working in the audio engineering field

Each of them

carried out the three sessions consecutively

  • 2 presents the listening room in use in this study

room is a recording studio for amplified music

  • listening position is at 1
  • 5 m from the nearest wall

same holds for all loudspeakers,

  • which is in agreement with

AES and IEC recommendations [1,

All of the four

loudspeakers are hidden behind a visually opaque,

  • acoustically transparent,

They are located at 2

from the centre of the listener’s head

The tweeters are

placed at the height of the listener’s ears

The distance

between two contiguous tweeters is 0

  • 5 m to keep

interactions between them as low as possible while

unaffecting the timbral quality

Each loudspeaker layout was tested by two different

  • listeners,

so as to present twice all of the 24 possible

  • combinations
  • 1 Preference scores

The range of the preference scale was continuous and

  • extended from
  • -1 to +1 depending on whether the

preference was marked for loudspeaker A or loudspeaker B,

  • respectively

The answer to each trial was thus a preference

  • probability within

-1 and 1

Since listeners were free to use

the answer scale at their convenience,

  • no normalization was
  • applied to the results

Let us denote Pij,

  • the preference
  • probability of stimulus,
  • versus stimulus,

It is assumed

As shown in Table 1,

the reverberation time for this room,

measured between 125 and 4000 Hz,

  • ranges between 0

which agrees with IEC specifications [2]

Results

Pij = − Pji

Table 1 Reverberation time measured by octave bands in

the listening room used for the tests

A negative probability of preference Pij means that the

loudspeaker j is preferred to i

The linear relation described

Acoustics 08 Paris

(3) allows one to derive the preference scores from

these preference probabilities:

Si = ∑ Pij

Preference score

where Si is the preference score of stimulus i

As four

loudspeakers were compared in this experiment,

preference score could theoretically lie within

-3 and +3

A multivariate analysis of variance was made to see

whether the preference scores were affected by the

  • loudspeaker,
  • the loudspeaker position,
  • the session and the
  • excerpt
  • 2 Loudspeaker effect

Loudspeaker location

The analysis of variance showed that the most influential

factor was the loudspeaker itself (F(3,1724) = 87

However,

the Fisher LSD test showed that,

  • among the loudspeakers,

only item A was significantly less

appreciated than the three other ones (p < 0

  • shows that loudspeakers B,

C and D'were systematically

  • preferred to loudspeaker A
  • 4 Mean preference scores for the four positions,
  • their 95% confidence interval
  • 4 Loudspeaker/Session interaction

The interaction between the loudspeaker and the session

was significant (F(6,1721) = 3

About the third session,

  • 5 clearly shows the occurrence

of more subtle preference ratings than the global scores

  • displayed on Fig

As previously observed for the

  • loudspeaker effect,

the only significant differences in

sessions 1 and 2 were between item A and the three other

  • ones (p < 0

In session 3,

  • the ratings about

loudspeaker B were significantly different from the ones for

  • loudspeakers C and D'(p < 0
  • 05 and p < 0
  • respectively)

It means that the third session (where the

listeners were allowed to adjust the reproduction level) was

more discriminating than the two other ones

Preference score

It is worth noting that the preference scores for a given

loudspeaker could significantly vary between two sessions

The scores obtained by Loudspeaker A during sessions 1

and 2 were significantly different (p < 0

Loudspeaker

  • 3 Mean preference scores for the four loudspeakers,

within their 95% confidence interval

Speaker A

Speaker B

Speaker C

Speaker D

  • 3 Location effect

Preference score

The loudspeaker location had also a significant influence on

the preference scores (F(3,1724) = 63

  • shows that,

when the loudspeakers were set in front of the

listener (positions 2 and 3 according to Fig

appraisal were significantly better than when they were on

  • the sides (positions 1 and 4)

This finding may result from

different reasons: i) the listening room excitation depends

  • on the loudspeaker location

ii) according to some listeners,

the loudspeaker identification could be easier when the

loudspeaker is set on the side (positions 1 and 4) and iii)

when a loudspeaker is placed away from the listener’s axis,

he may need to turn his head towards the source,

  • and this

movement could have a negative effect upon his sound

  • assessment

Session

  • 5 Mean preference scores for the four loudspeakers in
  • the three sessions,

within their 95% confidence interval

No interaction was found between the loudspeaker and its

location: the effect of the location was the same for the

  • different loudspeakers
  • 5 Loudspeaker/Excerpt interaction

A significant interaction was also found between the

loudspeaker and the excerpt (F(6,1721) = 8

Acoustics 08 Paris

Loudspeaker A got the lowest preference scores,

  • whatever
  • the excerpt

loudspeaker C was significantly preferred

  • to D'(p < 0
  • 05) for the 2nd excerpt (Ben Harper)

For the

  • third excerpt (Gershwin),
  • loudspeakers B and D'were

significantly preferred to loudspeaker C (respectively

  • 001 and p < 0

This excerpt was a piano solo

recording containing impulsive sounds

The piano was

  • recorded live,

and the recording was therefore a bit noisy

One should note that,

with the Bernstein extract (symphonic

orchestra with large masking effect),

  • the loudspeakers B,
  • and D'were equivalently rated

References

  • [1] Audio Engineering Society,

AES20-1996: AES

recommended practice for professional audio subjective evaluation of loudspeakers (reaffirmed

  • 2007)"

Journal of the Audio Engineering Society

  • 382-400 (1996)
  • [2] International Electrotechnical Commission,
  • system equipment
  • - Part 13: Listening tests on
  • loudspeakers"

IEC Publication 60268-13 (1998)

Subjective measurements of loudspeaker

sound quality and listener performance"

Journal of the

Audio Engineering Society 33(1/2),

  • 2-32 (1985)

Conclusion

Gabrielsson,

Rosenberg,

Sjögren,

Judgments

and dimension analyses of perceived sound quality of

sound-reproducing systems"

Journal of the Acoustical

Society of America 55(4),

  • 854–861 (1974)

This study dealt with the comparison of three different

procedures designed to assess loudspeaker sound quality

The listeners were proposed successive or alternate

  • listening,

where the reproduction level was matched or left

Successive and alternate presentation provided

equivalent results when the loudness was matched

given the possibility of setting the reproduction level,

listeners tended to give more discriminating assessments

Moreover the different tests showed significant effects of

the extract and the loudspeaker position,

  • which was already
  • shown in previous studies

Zwicker,

Psychoacoustics – Facts and

  • models"

Springer Verlag (1990)

Glasberg,

A model for

  • the prediction of thresholds,
  • loudness,
  • and partial
  • loudness"

Journal of the Audio Engineering Society

  • 224-240 (1997)

Schuck,,

Bonneville,

The effects of loudspeaker placement on listener

  • preference ratings"

Journal of the Audio Engineering

Society 42(9),

  • 651–669 (1994)

Acknowledgments

Lavandier,

Herzog,

Meunier,

Comparative

measurements of loudspeakers in a listening situation"

Journal of the Acoustical Society of America 123(1),

  • 77–87 (2008)

The authors are grateful to the staff and students from

  • ‘Image & Son Brest’,

Université de Bretagne Occidentale

A multiple regression model for

predicting loudspeaker preference using objective

  • measurements: Part 1
  • - Listening test results"

Proceedings of the 116th AES Convention (2004)

Subjective measurements of loudspeakers:

A comparison of stereo and mono listening"

Proceedings of the 74th AES Convention (1983)

Audio engineering

  • - Science in the service
  • of art"

in Proceedings of the 111th AES Convention

Design considerations for loudspeaker

  • preference experiments"

Journal of the Audio

Engineering Society 40(12),

  • 979–996 (1992)

Perception of timbre of reproduced sound in

small rooms: Influence of room and loudspeaker

  • position"

Journal of the Audio Engineering Society

  • 42(12),
  • 999–1007 (1994)

We use coockies Savoir plus Close