Recordings of Italian opera orchestra and soloists in a silent room

Abstract

Anechoic recordings of symphony orchestras have been proposed in the literature and have been used in a multitude of studies concerning both innovative measurements and psychoacoustic experiments. Using the same approach, the present work shows the results of a recording campaign focused on the Italian Opera. Different motifs from Italian Operas have been played by professional musicians and soloists in the silent room of the Bologna University. The excerpts have been chosen because of their musical style characteristics and their acoustic properties (dynamics, timbre,vibrato). The chosen motifs come from scores of Donizetti, Verdi and Puccini, in order to consider various orchestrations and Opera styles.

Full text

Recordings of Italian opera orchestra and soloists in a silent room
Dario D’Orazio, Simona De Cesaris, and Massimo Garai
Citation: Proc. Mtgs. Acoust. 28, 015014 (2016); doi: 10.1121/2.0000425
View online: http://dx.doi.org/10.1121/2.0000425
View Table of Contents: http://asa.scitation.org/toc/pma/28/1
Published by the Acoustical Society of America

Published by the Acoustical Society of America
Volume 28 http://acousticalsociety.org/
22nd International Congress on Acoustics
Acoustics for the 21st Century
Buenos Aires, Argentina
05-09 September 2016
Architectural Acoustics: Paper ICA2016 - 732
Recordings of Italian opera orchestra and soloists
in a silent room
Dario D’Orazio, Simona De Cesaris and Massimo Garai
DIN, University of Bologna, viale Risorgimento 2, 40136 Bologna, Italy; dario.dorazio@unibo.it;
simona.decesaris2@unibo.it; massimo.garai@unibo.it
Anechoic recordings of symphony orchestras have been proposed in the literature and have been used in a
multitude of studies concerning both innovative measurements and psychoacoustic experiments. Using
the same approach, the present work shows the results of a recording campaign focused on the Italian
Opera. Different motifs from Italian Operas have been played by professional musicians and soloists in
the silent room of the Bologna University. The excerpts have been chosen because of their musical style
characteristics and their acoustic properties (dynamics, timbre, vibrato). The chosen motifs come from
scores of Donizetti, Verdi and Puccini, in order to consider various orchestrations and Opera styles.
© 2017 Acoustical Society of America [DOI: 10.1121/2.0000425]
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 1

1 Introduction
Anechoic recordings have been in use since the ’70s for listening tests. The earlier anechoic excerpts
were recorded in the BBC anechoic chamber [1] and widely used, e.g in the development of Ando’s theory
[2]. They consist of compositions for a small orchestra, recorded by one point microphone, with a low
dynamics.
In 1988 Hidaka et al. [3] recorded an orchestra in a normally reverberant concert hall (the Minoo Civic
Hall in Osaka, Japan), but surrounding the stage with an acoustically absorptive enclosure. The study was
mainly oriented to compare different miking techniques. A large orchestra was used: the string section was
composed by 16 first violins, 13 second violins, 11 violas, 8 cellos and 6 double basses, 4 musicians for
each woodwind instrument, a large brass section, 4 percussions, etc. . . 22 microphones have been used at
the same time during the performance: 8 used as “main” and the remaining as “spot” microphones for each
section.
In the same years, in the frame of the ARCHIMEDE project [4], supported by Bang & Olufsen and DTU,
excerpts for soloist (speakers and musicians) in various acoustic conditions (anechoic chamber, listening
room, studio, church) were recorded.
While in the ARCHIMEDE project solo performances were recorded, Vigeant et al. [5] recorded sym-
phonic music using multichannel techniques. Musicians were recorded individually in the DTU anechoic
room, using a video recording of the conductor and a 3D array of five omnidirectional microphones. Two or
three musicians have been recorded for each string part, one musician for the remaining parts. The record-
ings have been used by musicians as a promotional tool and by researchers as a sound source for MIMO
auralizations [6].
P¨
atynen et al. [7] recorded symphonic and opera music by using an approach similar to the Vigeant’s
one, but extending the miking spatial resolution. Several musicians and a soloist were recorded individually
in the anechoic room of Espoo University, using an array of 22 microphones.
Other projects provided anechoic recordings of soloists and choir [8, 9, 10].
2 Selecting the music material
The present work aims at extending the availability of recorded (quasi)–anechoic material, introducing
three Italian opera recordings, played by orchestra and soloists. Music material for the present recordings
was chosen to represent the most performed Italian composers and their styles.
Under the definition of “Italian opera” there are compositions from the XVII century (the birth of melo-
dramma, G. B. Peri, etc. . . ) to the contemporary years. Each period has been characterized by a writing
style, from early polyphony to serialism. At the same manner in each period a listening ambient has been
used, from the first court rooms (e.g. the Cornero Odeum in Padua) to the multipurpose opera houses of
the XXI century. Finally the compositions of various ages need different orchestral compositions: few
instruments and basso continuo for the recitative in the XVIII century, a large orchestra in Puccini’s operas.
The statistics of opera performances in the 2015–2016 season (see Annex A for details), shows that
Italian authors and Mozart are dominant (Figs. 1 and 2). These results agree with Hidaka’s ones [11], in
which the 1997–98 season at 32 major opera houses around the world was analyzed. Basing on a statistical
results, the music material for the present study has been chosen from the classical age of the melodramma,
which spans from the early XIX century to 1930s.
The first excerpt is an aria (“Come Paride vezzoso”) from the opera L’elisir d’amore by Gaetano
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 2

Table 1: Synthesis of the previous anechoic recordings and the presented ones.
Author(s) Type Mics Location Music materials Year
Burd [1] ensemble 1 BBC
Gibbon: Royal pavane
1969
Mozart: Symphony KV 551, 4th mov
Arnold: Sinfonietta, opus 48, 4th mov
Haydn: Symphony no.102, 2nd mov
Wagner: Siegfried Idyll
Hidaka et al. [3] ensemble 29 Theatre1
Haendel: Water music
1988
Mozart, Le Nozze De Figaro (Ouv.)
Beethoven: Symphony no.3, 4th mov
Glinka: Ruslan And Lyudmila (Ouv.)
Verdi: La Traviata (Preludio)
Brahms: Symphony no.4, 1st mov
J. Strauss: Pizzicate-Polka
Bizet: L’Arlesienne, Minuet
Bruckner: Symphony no.4, 1st mov
Debussy: Apres-midi d’un faune
Mahler: Symphony no.5, 4th mov
Mussorgsky: Pictures at an exhibition
ˇ
Sostakoviˇ
c: Symphony no.5, 1st mov
Hansen and Munch [4] solo 1 DTU excerpts for solo21991
Vigeant [5] solo35 DTU
Brahms: Symphony no.4, 3rd mov
2005Mozart: Symphony no.40, 1st mov
Stravinsky: Circus Polka
P¨
atynen et al. [7] solo 22 Espoo
Mozart: Don Giovanni, aria
2009
Beethoven: Symphony no.7, 1st mov
Brukner: Symphony no.8, 2nd mov
Mahler: Symphony no.1, 4td mov
(1) Minoo Civic Hall with “anechoic room installed around stage”; (2) Excerpts for guitar, cello, percussions, trumpet and cornet
solo; (3) music materials wasn’t available
Donizetti (1797-1848), whose first representation was in 1832 at the Teatro della Cannobiana of Milan
(now Teatro Gaber). The excerpt represents the belcanto in the Italian opera: a cavatina for coloritura bari-
tone in which the figure of Belcore appears in the first act. The soloist part includes a cadenza and some
agility passages (see Fig. 3). In this motif there are various soloists parts (the tenor Nemorino, the sopranos
Adina and Giannetta) and a choir. The score shows several tempo variations: the “larghetto” in 3/4 then the
“andantino” in 4/4, the free “colla voce”, which resolves to the initial time (“a tempo”) in the final.
The second recording is extracted from the opera Il trovatore by Giuseppe Verdi (1813-1901), first
represented in 1852 at Teatro Apollo of Rome (now demolished). In the cabaletta of Leonora “Di tale amor,
che dirsi” the soloist figure is a lyric soprano in which voice weight and agility coexist. In this excerpt the
Verdian orchestra is complete (strings, full woodwinds with ottavino, full brasses with bass trombone). The
final part presents a crescendo with a wide dynamic range and an accelerated tempo. It is worth noting that
the successful result of an opera singer is due also to facial expression and body movements, which add
something to the singing voice but make very difficult to fix it in a single audio record.
In order to complete the temporal evolution of the Italian Opera, the third motif is the romanza “Oh Mio
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 3

!"##$%$&
'(&
)*++$%$&
,(& -"%$.*/&
0(&
1*23$&
,4(&
5$"236%"&
,(&
7899$%$&
:(&
)**;<"=*%&
,(&
>8?@*23$%9A&
,(&
B6C%*2&
:(&
!DE;268##&
0(&
F".62;&
,G(&
)$.*;&
H(&
F6##*%*;&
,(&
)*2+$".&
,(&
5"8%"3&
,(&
F8#"2C#A$I&
,(&
-="JKA&
,(&
L6IA"=#A$I&
4(&
7829*++&
,(&
)2$M*%&
,(&
>N%3*+&
,(&
O6%KP*A&
,(&
Q;<*2#&
H,(&
Figure 1: Analysis of the authors and operas represented in the 2015–2016 season at 51 of the most important
opera houses (see Annex A for details): the most represented composers (green: Italian authors, yellow:
German authors, red: Mozart, blue: French authors, etc...).
Table 2: Presented anechoic recordings.
Composer opera Act Aria
Donizetti L’elisir d’amore 1st Come Paride vezzoso
Verdi Trovatore 2nd Di tale amor, che dirsi
Puccini Gianni Schicchi 1st O mio babbino caro
Babbino Caro” from the opera Gianni Schicchi by Giacomo Puccini (1858–1924), whose first representation
was at Metropolitan Theater of New York City in 1918. Here the soprano sings with slight voice and the
music is a siciliana played by a large string section, few woodwinds and a harp. The recording includes
initial clusters for strings and brasses.
A fourth excerpt (the aria “Tacea la notte placida” from Verdi’s Trovatore) was discarded after the
recordings, due to problem of mismatched timing between sections. On one hand the music accompaniment
is written as regular sequence of triplets. On the other hand the execution needs a continuous variation in
tempo and dynamics, in order to emphasize the soloist. When the metronome is quite low (“andante” in the
present case) there are several problems of mismatching timing when each instrument is recorded as solo.
3 The orchestra
Professional musicians from the Corelli Orchestra of Ravenna and soloists were asked to join this re-
search. Some musician already had experiences in recording contemporary and pop music too. The ar-
rangement of the recording has been similar to the one in previous literature [5, 7]: only one musician
per instrument played all parts one after another, following a reference video of the conductor with a pi-
anist. During the recording takes, the musicians heard the sync piano and the previous tracks through closed
headphones. Following musician requirements, the right channel only was powered to the headphones for
self-hearing.
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 4

!"
#"
$!"
$#"
%!"
%#"
&'()*+,"-./"0)12/*34+/"
&'()*+,"5/"6'((/"7."89)*'"
:1;;.6.,"5)<'=>?/"
:1;;.6.,"@'A;)"
B/*7.,"5) @*)C.)+)"
<.(/+,"D)*?/6"
B/*7.,"E.9'F/G'"
E'AA.6.,"HF"<)*2./*/"7."I.C.9F.)"
&'()*+,"-'6"J.'C)66."
:1;;.6.,"&)7)?)"<1G/*3K"
L)96/*,"-./"L)FMN*/"
EO"I+*)1AA,"I)F'?>"
B/*7.,"P)21;;'"
L)96/*,"-/*"3./9/67/"Q'FFR67/*"
Figure 2: Analysis of the authors and operas represented in the 2015–2016 season at 51 of the most important
opera houses (see Annex A for details): the most represented operas. The colors point out the language of
the operas: German (yellow), Italian (green), French (blue).
Table 3: Orchestral parts during recordings (vli: violin, vla: viola, c: cello, db: double-bass, fl: flute, ob:
oboe, cla: clarinet, bas: basson, ho: horns, tba: trumpet, tbn: trombone, ha: harp).
Excerpt 1st vli 2nd vli vla c db fl ob cla bas ho tp tbn ha
Donizetti 8 6 5 4 3 2 2 2 2 2 2 3 –
Verdi 10 8 6 6 4 2a2 2 2 4 2 4b–
Puccini 12 10 8 7 5 2a3c3d2 4 3 – 1
(a) a flute and a piccolo; (b) three trombones and a bass trombone; (c) a oboe and a English horn; (d) two clarinets and a bass
clarinet.
Three soloists and thirteen musicians were recorded. With respect to P¨
atynen’s work, where each string
part has been recorded once, in the present work the strings have been recorded several times (see Tab. 3).
The conductor attended all the recording sessions, in order to evaluate the effectiveness of the single takes
on the general impression of the orchestra. Double basses and cellos have been recorded in the first session,
followed from the second violins and the violas. Only a single track of the first violin has been recorded
as reference for the intonation. In the second session brasses have been recorded: horns, trombones and
trumpets. From the third session (woodwinds) onward the piano track has been removed from the audio
monitoring heard by musicians during recordings. In the fourth session all the parts of the first violins have
been recorded, looking for the proper ’color’ of the orchestra. In the last session a harp was recorded and
some takes were overdubbed.
The recording room was the silent room of the University of Bologna. The room was measured using a
prototype of omnidirectional sound source [15] and the extracting the reverberation time using the envelope
of the IRs [13, 14], due the high slope of the decay. The faint reverberation at low frequencies (see Fig.
4) was not considered as a problem, since the decay time of the instruments which produce fundamental
frequencies below 150 Hz.
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 5

Figure 3: Score for soloist and piano accompaniment of the final of the first Donizetti’s excerpt: agilities of
the soloist on the larghetto reprise and final cadenza (“A–mor”).
4 Recording techniques
A great deal of attention has been devoted in the previous studies to the miking techniques. Hidaka et
al. used several small diaphragm mics: B&K 4003 (omni), B&K 4006 (omni) and Schoeps for the main, as
a spot Schoeps for woodwinds and strings, Neumann SM-81 for brasses and AKG C451E for percussions.
Hansen and Munch used both cardioid (Sennheiser MKH40, B&K 4011, Schoeps MK4) and omnidirec-
tional (B&K 4003) mics, depending on the environment. Vigeant et al. used small diaphragm DPA 4006
omnidirectional microphones. P ¨
atynen et al. used large diaphragm Rode NT1 [7] microphones.
Audio-Technica AT 4050 large diaphragm microphones have been used in the dodecahedrical array used
in the present work, thanks to the good recording capability and low noise characteristics (see Fig. 5). The
microphones have been used in their omnidirectional configuration and have been preamplified and AD
converted by a RME Micstasy, set with about 35 dB of gain. A pad attenuation of -15 dB was used for the
trombone takes only.
Moreover, several reference microphones have been placed in the room in order to compare and equalize
the recorded tracks (see Tab. 4 for details of the configuration).
5 External resources
Some of previous anechoic works have been distributed in a commercial CD [20, 21] or free download
[7, 8, 9]. The recorded material of Vigeant’s work [5, 6] wasn’t distributed.
Audio tracks of this work (recorded at 48 kHz/24 bit) are freely available for academic uses. See more
at http://acustica.ing.unibo.it/opera.
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 6

63 250 1,000 4,001
0
0.1
0.2
third-octave bands (Hz)
T15 (s)
Figure 4: Reverberation time measured in the silent room during the recordings.
Table 4: Details of the microphone configuration used during the recordings.
No. Type Elevation (degree) Azimuth (degree) r (m)
1 AT 4050 52.6 120 1.1
2 AT 4050 52.6 0 1.1
3 AT 4050 52.6 240 1.1
4 AT 4050 -10.8 240 1.1
5 AT 4050 10.8 300 1.1
6 AT 4050 -10.8 0 1.1
7 AT 4050 10.8 60 1.1
8 AT 4050 -10.8 120 1.1
9 AT 4050 10.8 180 1.1
10 AT 4050 -52.6 60 1.1
11 AT 4050 -52.6 180 1.1
12 AT 4050 -52.6 300 1.1
13 B&K 4190 -6,5 107 2.7
14 B&K 4190 -15.3 0 1.1
15 B&K 4190 -15.3 120 1.1
16 AT 4050 -6.5 103 2.7
17 AT 4050 -8 105 2.2
6 Conclusions
Quasi-anechoic recordings of opera excerpt were presented. Soloists and musicians were recorded fol-
lowing procedures similar to those found the previous literature. All string parts were recorded using a
complete string section, playing one instrument at a time.
The recorded material is free of use for academic uses. In the authors’ hope, the excerpts may be useful
for further researches in the field of opera house acoustics, extending the available corpus of anechoic
recordings [22].
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 7

Figure 5: Recording setup during the recordings.
Acknowledgements
The authors gratefully acknowledge the musical contribution of the conductor Jacopo Rivani and the
soloists and the musicians of the Corelli Orchestra, expecially Nicol´
o Grassi and Simone Marzocchi. The
microphones were provided by Andrea Guerranti of SISME SpA.
References
[1] A. N. Burd (1969). Nachhallfreie Musik fur Akustische Modelluntersuchungen. Mitteilungen, 13, 200–
201.
[2] D. D’Orazio, S. De Cesaris and M. Garai (2011). A comparison of methods to compute the “effective
duration” of the autocorrelation function and an alternative proposal, J. Acoust. Soc. Am. 130(4), 1954–
1961.
[3] T. Hidaka, K. Kageyama, S. Masuda (1988), Recording of anechoic orchestral music and measurement
of its physical characteristics based on the auto-correlation function, Acustica, 67 68–70.
[4] V. Hansen, G. Munch (1991). Making recordings for simulation tests in the Archimedes project, J.
Audio Eng. Soc., 39(10), 1991 October.
[5] M. C. Vigeant, L. M. Wang, J. H. Rindel, C. L. Christensen, and A. C. Gade (2010). Multi-Channel
Orchestral Anechoic Recordings for Auralizations, Proc. of the International Symposium on Room
Acoustics (ISRA 2010), Melbourne, Australia, August 2010.
[6] M. C. Vigeant, L. M. Wang, and J. H. Rindel (2008). Investigations of orchestra auralizations using the
multi-channel multisource auralization technique, Acta Acust. united Ac., 94, 866–882.
[7] J. P¨
atynen, V. Pulkki, and T. Lokki (2008), Anechoic recording system for symphony orchestra, Acta
Acust. united Ac., 94(6), 856–865, November/December 2008.
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 8

[8] Z. Karabiber (2000). A new approach to an ancient subject: CHARISMA Project, in 70th International
Congress on Sound and Vibration, Garmisch-Parterkirchen, Germany, 4–7 July 2000.
[9] Z. Yuksel, S. Erdogan (2005). Virtual conservation of acoustical heritage: CHARISMA and ERATO
Projects, Proc. Forum Acusticum 2005, Budapest.
[10] R. Freiheit (2010). Creating an Anechoic Choral Recording, Proceedings of the International Sympo-
sium on Room Acoustics, ISRA 2010, Melbourne, Australia, 29-31 August 2010.
[11] T. Hidaka and L. L. Beranek (2000). Objective and subjective evaluations of twenty-three opera houses
in Europe, Japan, and the Americas, J. Acoust. Soc. Am., 107 (1), 368–383.
[12] T. Lokki (2007). How many point sources is needed to represent strings in auralization?, International
Symposium on Room Acoustics (ISRA 2007), Seville, Spain, September 10–12 2007. Paper P11.
[13] D. D’Orazio, S. De Cesaris, M. Garai (2012). Measuring reverberation time using preprocessed energy
detection, Internoise 2012, New York City, August 2012.
[14] S. De Cesaris, D. D’Orazio, F. Morandi, M. Garai (2015). Extraction of the envelope from impulse
responses using pre-processed energy detection, J. Acoust. Soc. Am., 138, 2513.
[15] D. D’Orazio, S. De Cesaris, P. Guidorzi, L. Barbaresi, M. Garai, R. Magalotti (2016). Room acoustic
measurements using a high SPL dodecahedron, 140th AES Convention, Paris, May 2016.
[16] AT4050 datasheet. http://www.audio-technica.com (last viewed November 28, 2016)
[17] B&K 4190 datasheet. http://www.bksv.com/ (last viewed November 28, 2016)
[18] Reaper v. 5.20 http://www.reaper.fm (last viewed November 28, 2016)
[19] M. Garai, F. Morandi, D. D’Orazio, S. De Cesaris, L. Loreti (2015). Acoustic measurements in eleven
Italian opera houses: Correlations between room criteria and considerations on the local evolution of a
typology, Building and Environment, 94(2), 2015, 900–912.
[20] Bang & Olufsen: Music from Archimedes. Audio CD 101, 1992.
[21] Denon: Anechoic orchestral music recording. Audio CD, Denon Records, 1995. ASIN: B0000034M9.
[22] A. Kuusinen (2014). An anechoic audio corpus for room acoustics and related studies, Proc. of the
EAA Joint Symposium on Auralization and Ambisonics, Berlin, Germany, April 2014.
Appendix A
The opera performed in the 2015–2016 season were analyzed in the following theatres: Aalto Musik-
theater Essen, Bayreuther Festspielhaus, Bayerische Staatsoper (M¨
unchen), Bolshoi Theater (Moscow),
Bremen Theatre, De Nationale Opera (Oslo), Deutsche Oper (Berlin), Erkel Theatre (Budapest), Glynde-
bourne Opera House, Gross Festspielhaus Salzburg, La Monnaye (Bruxelles), Leipzigoper, Lyceum The-
ater (Barcelona), Luzerner Theater, Magyar ´
Allami Operah´
az (Budapest), Metropolitan Theater (New York
City), Nationaltheather Mannheim, New National Theatre (Tokyo), Op´
era Bastille (Paris), Odeon Theatre
Nancy, Ooperabaletti (Helsinki), Op´
era Grand Avignon, Op´
era National de Bordeaux, Op´
era National du
Rhin (Strasbourg), Opera Royal de Li´
ege, Opernhaus Z¨
urich, Oper Frankfurt, Oper Stuttgart, Palais Gar-
nier (Paris), Royal Danish Theater (Copenhagen), Royal Opera House (London), Seattle Opera, Semper-
oper Dresden, Staatsoper Hamburg, St´
atn´
ı Opera (Prague), Sydney Opera House, Teatro alla Scala (Milan),
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 9

Teatro Colon (Buenos Aires), Teatro Campoamor (Oviedo), Teatro Carlo Felice (Genova), Teatro Comu-
nale (Bologna), Teatro dell’Opera (Rome), Teatro La Fenice (Venice), Teatro Real (Madrid), Teatro Regio
(Parma), Teatro Regio (Torino), Teatro San Carlo (Naples), Teatr Wielki (Warsaw), Tokyo Opera City Con-
cert Hall, War Imperial Opera (San Francisco), Washington National Opera, Wiener Staatsoper. Operettas
and musicals weren’t taken into account in the statistics.
D. D'Orazio et al.
Italian opera recordings
Proceedings of Meetings on Acoustics, Vol. 28, 015014 (2017)
Page 10