Performance Practice
of Electroacoustic Music

Sotto Voce was commissioned by IRCAM and was premiered on 30 June 1995 by the British vocal ensemble “Electric Phoenix” with Judith Rees, soprano, Meriel Dickinson, mezzo, Daryl Runswick, tenor, Terry Edwards, bass and John Whiting, sound engineer. The electronics were realised at IRCAM with Serge Lemouton. The piece is dedicated to the Spanish sculptor Francisco Luque (*1948).
The composition was inspired by Castilian scholar Pedro Alfonso de Huesca’s (1062?–1140?) text “De silentio” from his Disciplina clericalis, a collection of 33 exempla; observations and advice offered by a philosopher to his disciples. To the composer the text is a lesson in modesty and thoughtfulness, though not without some ironic moments. His main interest concerns however the text’s underlying expressive potential and particularly its phonetic structure. [Interview, 3:40–7:00 and 58:10–1:01:50]

De silentio

[…] Ait enim philosophus: Silentium
est signum sapientiae, et loquacitas est signum stultitae. [- Alius:] Ne festines respondere
donec fuerit finis interrogationis, nec quaestionem in conventu factam solvere temptes, cum
sapientiorem te ibi esse prospexeris, nec quaestioni alii cuiquam factae respondeas, nec
laudem appetas pro re tibi incognita.

Philosophus enim dicit: Qui de re sibi ignota laudem
appetit, illum mendacem probatio reddit. [- Alius:] Adquiesce veritati sive a te prolatae sive
tibi obiectae.

[- Alius:] Ne glorieris in sapientibus verbis tuis, quia prout philosophus testatur:
Qui in suis verbis sapientibus gloriatur, stultus esse comprobatur. – Haec omnia faciens
connumeraberis inter discipulos sapientiae atque prudentiae.
Philosophus dicit: […]

[…] For the philosopher says: Silence is a sign of wisdom, and verbosity is a sign of stupidity. [- Another:] Do not hasten to answer before the question has been finished, and do not try to answer a question raised in a meeting, if you see someone wiser than you there. And do not respond to a question that was put to someone else. Also, do not seek praise for something unknown to you.

For the Philosopher says: He who seeks praise for someone unknown to him will be exposed as a liar in the exam [- Another] assent to truth, whether because you studied it, or because it lies before you.

[- Another:] Do not boast with your words of wisdom, because, as the philosopher testifies: He who boasts about his words of wisdom proves to be a fool. – If you practice all of this, you will be counted among the disciples of wisdom and of prudence.
The philosopher says: […]

The sound material of Sotto Voce was developed in close collaboration with the performers:

“Un premier pas dans la composition de l’œuvre a consisté à écrire et tout de suite enregistrer – par les mêmes chanteurs qui vont créer cette pièce – un ensemble de figures, gestes, paroles et phonèmes tirés du texte […]” [López López, Note de programme]

The collaboration entailed both the recording of pre-composed vocal sequences and the exploration of vocal techniques for generating materials, which were then electronically processed. [Interview, 10:54 and 13:40–15:00]
Sotto Voce testifies to López López’ early interest in working with different temporal levels, seeking to explore their expressive potential and to create relationships between micro and macro structures. According to López-López temporal levels extend from the infinitesimal to the cosmic time, the single note being at the border between the micro- and the macro-temporal domain. In such a wholistic view the specific musical time domains are only a part of a far more complex layering system. It includes e.g. the paradoxical situation of experiencing events in the present that originated millions of years ago. In fact, phenomena from physics and natural sciences constitute a constant source of reflection for López López. [Interview, 22:38]
Of particular interest is the development of “micro figures”. In Sotto Voce micro figures are small samples containing e.g. gestures or rhythmic cells. They are processed and projected onto various time scales in order to create complex polyphonic textures and generate new micro figures that can then be treated in a similar way. [Interview, 20:00–23:20]. This process, related in particular to ideas of Horacio Vaggione and to spectralist theory, [Interview, 27:47 and 35:35–37:10] is likewise meant to define temporal relationships and to create transitions between harmony and timbre. [Interview 33:30–34:45].
In the electronic part, such spectral transformations are achieved through analysis and re-synthesis of vocal material:

“Un niveau compositionnel complémentaire s’ajoute au précédent grâce aux techniques de re-synthèse d’un son permettant de travailler à l’intérieur de la voix, c’est-à-dire la reconstitution numérique de la voix grâce à l’analyse et à la transformation des paramètres qui la constituent au moyen des oscillateurs électroniques. [López López, Note de programme]

These techniques allow to process acoustic material even to the point of resembling purely electronic sounds [12:06] that are then used to generate various kinds of “vocal synthesis”. This approach resembles vocal-electronic relationships in pieces such as Mortuos Plango, Vivos Voco (1980) by Jonathan Harvey and Les Chants de l’Amour (1984) by Gérard Grisey (Links zur Database) and continued to play an important role in López López’ subsequent musical production.
The harmonic structure of Sotto Voce is thus largely determined by fields of spectral chords, as can be observed right after the introduction. These are based on a 12-tone row containing the even harmonics of F# up to the 23th. Vocal passages clearly related to the fundamental gradually shift towards disharmonic spectra, creating points of transition to textures of “electronic” quality. This involves both live and electronically processed voices with extended vocal techniques such as glissandi, trills and whistles. Noise-like textures based on single phonemes and vocal articulations or containing interjections, spoken and whispered words and sentences create a sonic counterpoint conveying some of the semantic content of the text. The multilayer structure is enhanced by the quadraphonic disposition of the tape as visualised in the score.
The piece is also characterised by a close relation between the sung material and pre-produced sounds. Its realisation will be obviously challenging due to differences between the specific vocal qualities of the live and the original performers contained in the electronic part. This raises likewise questions with regard to amplification, intonation, pronunciation and the articulation of phonemes. The composer advises performers to mainly orient themselves by following the tape, allowing, however, some margin for differences and nuances. [Interview, 42:00–43:20]
In sum, Sotto Voce displays a commitment to integrating different formal approaches and layers of musical textures rather than aiming for formal reduction or stylistic “purism”. In the words of López López „reunir más que filtrar.” (“to gather, rather than to filter out”). [Interview, 29:02]

1) Performance Materials:

a) Score:
-Editions Ricordi Paris (R2728), copyright 1995
-Editions Ricordi Paris (R2728), edition dated 27 October 2017

-The two editions are identical in content.

The score contains information on the placement of the singers and the speaker setup and schema of the signal flow.
The live vocal parts are written in traditional notation with a constant tempo (quarter note = MM 60).
The electronic part is notated using screenshots of the DAW session. It contains up to 13 audio tracks displaying the wave forms and the names of the single sound files. The timeline in seconds is included on top of the score to facilitate synchronisation.

b) Other materials:

Electronic part

1)
Source: Publisher/Studio CDN
Date: 23 November 2017 (files)
Format: Wav 48 kHz 32b
Synchronisation: Click track

Files:
Audio 1_Lopez_Lopez_Sottovoce_48khz_32bits.wav
Audio 2_Lopez_Lopez_Sottovoce_48khz_32bits.wav
Audio 3_Lopez_Lopez_Sottovoce_48khz_32bits.wav
Audio 4_Lopez_Lopez_Sottovoce_48khz_32bits.wav
Audio 5_Lopez_Lopez_Sottovoce_48khz_32bits.wav
Audio 6_Lopez_Lopez_Sottovoce_48khz_32bits_ClickTrack.wav

2)
Source: IRCAM
Date: N/A
Format: Aiff 44.1 kHz 24b
Synchronisation: Time code

Files:
CD Image: SottoVoce_Pistes_Separees.dmg, containing:
Sottovoce 1.aif
Sottovoce 2.aif
Sottovoce 3.aif
Sottovoce 4.aif
Sottovoce 5.aif
Sottovoce TC.aif

3)
Source: Composer
(undated)
ADAT Tape 48 kHz 24b
Synchronisation: Timecode
Remarks:
Copy of original master
Includes indication for fifth channel to be distributed to channels 1-4

4)
=No. 3, digitised at ICST
Date: 18 February 2020 (digitisation)
Format: Wav 48kHz/24b
Synchronisation: Timecode

Files:
Wav 48kHz 24b
01-200211_ICST_48Khz24b.wav
02-200211_ICST_48Khz24b.wav
03-200211_ICST_48Khz24b.wav
04-200211_ICST_48Khz24b.wav
05-200211_ICST_48Khz24b.wav
08-200211_ICST_48Khz24b.wav

Editorial Instructions

Disposition
The score provides a stage plot and a signal flow plot, cf. above. The four singers are placed in the middle of the space, a microphone is provided for each. Four loudspeakers are to be placed at the corners of the hall. The audience is seated around the singers.

Synchronisation
A timecode track is included to facilitate synchronisation between the tape and the singers.
There are five audio tracks, one for each of the four speakers and an additional fifth track.. The latter should be routed to all four speakers; however, this information was found only on the label of the ADAT master tape (cf. above, Sources).

Fig. I. Sotto Voce, score.

Remarks by the composer (cf. Interview)

Amplification
Amplification should be carefully applied in order to achieve the best possible fusion between live voices and tape. [42:58]

Reverb
Reverb may be used if necessary; here, too, the fusion of both layers remains the objective. [44:38] The technician of “Electric Phoenix” most probably applied reverb [45:35].

Synchronisation
The click-track is intended for use during rehearsals only. In concert, “Electric Phoenix” used a visual clock instead. [46:23]

Sound projection
Slight level adjustments of the tape in relation to the live voices can be made live but must not affect the spatial relationships. [49:45] Subwoofer should be used very carefully. At the end of the piece the level can be slightly raised. [54:50]

Intonation and articulation
In general, the singers should follow the electronic part. However, differences are possible. [42:00] (See also performers’ remarks below)

Performance report

The concert took place on 17 January 2020 at the ZHdK’s concert hall 1 (KS 1). Performers were members of the Basle-based ensemble “SoloVoices” (Svea Schildknecht, soprano, Francisca Näf, alto, Jean Knutti, tenor and Jean-Christophe Groffe, bass), Leandro Gianini, sound engineering and Germán Toro Pérez, sound projection.

Tape issues
The tape provided by the publisher (cf. above, Sources) had some audible glitches. An ADAT tape with a digital copy of the original master was obtained from the composer, which was transferred to hard disc and compared to the other tape. Inverting the phase of one of the versions and recording the sum of the two, a perfect cancellation was achieved most of the time, so that only the glitches would be audible and visible in the resulting wave form. It can thus be said that the glitches were not present on the original master tape and most probably occurred in a subsequent transfer.

Position of the singers
Contrary to the instruction in the score, the singers were positioned in front of the audience. This setup proved more effective for projection of the live voices, resulting in a compact group sound. In the original disposition people in the audience would mainly hear the singers facing them. During the rehearsals, the composer expressed his general preference for a frontal disposition.

Amplification and sound projection
As stated above, the voices should be slightly amplified in order to obtain a homogeneous blend with the electronic part. Four Shure beta 53 headsets were used for the singers. The headsets should allow the singers to move freely and to obtain a constant microphone-source distance.
Six L-Acoustics 8XT speakers were used, four placed in the corners of the hall for playback of the tape and two additional speakers on the sides for reverb. [s. below] The singers were only amplified through the two front speakers. Additionally, one Meyer Sound 700-HP subwoofer was located in front. The fifth audio channel was routed to the four main speakers and to the subwoofer.

Fig. II. Routing

The sound movement is already contained in the tape; no other movement should be introduced from the FOH [Interview, 51:00– 53:00]
Three level master faders were set up for the control of the tape, the amplification and the subwoofers. To check the levels between the live ensemble and the electronic the section starting at 1:12 was used as reference. During the performance the levels were slightly adjusted.

Reverberation
As reverb unit, a Bricasti M7 was used (Settings: Mechanical Hall, decay time 2.2s, pre-delay 42ms.).
Only the live singers were reverberated. No reverberation was added to the tape. The reverb was routed to the six speakers. The two additional speakers added at the sides were intended to obtain a more uniform distribution of the reverb signal in the hall. The following table shows the corresponding levels and delay times.

Fig. III. Reverberation

Synchronisation
The timecode format is 25frs. Itt starts at 00:59:50:00. There is a 10 second count-in, the score start is at 0:00:00:00.
The tape is accompanied by a click track, however, as stated above, the composer strongly advises against using a click track during the concert, believing that the use of a timecode allows for a better connection with the tape sounds. “Solo Voices” used a time display (s. below), and one of the singers conducted intermittently.

Remark concerning the beginning
In the recording by “Electric Phoenix”, the piece starts with the sentence “Philosophus dicit silencium est” (“the philosopher says: silence is”) declaimed by the bass and followed by a noise-like sound texture. This sentence is not written in the live vocal part. According to the score, the tape starts at 0:00 with four sound files named “B Silencium est”. The reader thus assumes that these files contain that sentence. However, they are not present in the tape, and the first sound (the noise-like texture named “B cascadas fract.R” on track three) appears only at 0:02:096. The bass must declaim the missing sentence and synchronise it with the tape, so that the first sound file follows immediately as an answer.

Monitors
To facilitate synchronisation, we used two video monitors to display the timecode (one for two singers). The standard time code display and most of the DAW internal clock displays also show milliseconds or frames. This proved tiring for the eye, so that we would show only minutes and seconds. When using Reaper as a DAW we recommend a JS plugin ht.clock.min.sec provided to us by Hans Tutschku in order to achieve this.

Intonation and vocal articulation
One of the main issues during the rehearsal process concerned the tuning between the live singers and the tape. Some pitches in the tape weren’t perfectly in tune with corresponding notes in the score, for example:

-2’21 Alto H is too high
-7’40’’ Soprano/Alto C is too low
-7’50’’ Bass C sharp is too low
The singers tried to match the tuning to the tape as closely as possible.

Another issue is the pronunciation of the text. The score doesn’t specify this. The ensemble decided to use Italian Latin pronunciation.

Performance report by Ensemble “SoloVoices” (Edited by Svea Schildknecht)

1. Remarks concerning the material (score and tape)

The layout of the music
An edition with a larger text font, not handwritten but typewritten, would be desirable. A purely vocal score without electronics with the time indications placed directly above the sung lines would provide more clarity. The same applies to the placement of text syllables, dynamic indications, glissando indications, etc. below the respective vocal line.

2. Remarks concerning the execution

Conducting
We found it convenient for one of us to beat the time written by López López. Realising the rhythm precisely only reading the time line is difficult, as the eyes will focus on the music and not on the running clock. The conductor’s movements, on the other hand, can be seen from the corner of the eye.
Jean Knutti conducted four beats ahead of time 0:00 for the synchronous beginning with the playback tape, during which Jean-Christophe Groffe recited the sentence “Philosophus dicit:”. The score shows a discrepancy with the beginning of the sentence spoken by the bass.
Knutti gave signs for the changing of the lines in 0:23 and 0:47.
Before 1:12, Knutti again conducted four beats.

Pronunciation
A question arose as to the desired pronunciation of the Latin text: Italian, Spanish? We agreed on Italian, adapting the spelling for that purpose. We replaced the /c/ in the words “silencium”, “sapiencie”, etc. with a /t/.

Notation of text syllables and phonemes
There was some ambiguity with respect to the distribution of individual sounds to notes, such as at 1:51 in the soprano: si-g-num. The consonant cannot be sung to the length of an eighth note. It was not clear which phoneme the composer wanted. When asked, López López clarified that the should be added to the following syllable ‘-num’.
Even at 1:56 in the soprano line, and from 2:01 and 8:46 in all voices, the unvoiced of the word “stultitie” cannot be sung to the length of a whole note. López López changed the unvoiced to a voiced , and so the sound can now be sung as a phoneme [z].
The notation of the consonant at 2:19 looks ambiguous because it is graphically offset. We agreed on the synchronous completion on an eighth note break.
The ‘RS’ in 11:26 – 11:35 and 11:47 – 12:03 should be articulated as a flutter tongue with a clearly audible simultaneous [s] portion, i.e. with a wider body of the tongue.
The downward glissando in 13:19 – 13:35 on [s] can be made audible with the help of the vowel form from light [i] to dark [u].

Breathing while whistling
Breathing points for whistling at 3:53 – 4:45 were marked by López López as follows: with interruption of the glissando line, breaks and caesuras. The hooks in the curved line from 4:28 were also interpreted by us as breathing marks. Faster rebreathing can be achieved by switching from exhaled whistling to inhaled whistling.

3. Remarks concerning the interpretation – instructions from the composer and Germán Toro Pérez

The Tape
The tape is to be listened to actively by the singers and sung along to. There are individual passages with intonational ambiguities in the tape, such as 7:28 – 7:35. The performers were faced with the decision of adapting to the singers’ intonation on the tape or to keep a pure intonation. We chose to adapt to the tape in favour of homogeneity.

Volume and Balance
The volume levels of each singing line should always be clearly realised.
The same applies to accents such as in 8:36 – 8:43.
The balance of the live singing must be thoroughly balanced with the volume level of the tape.
The whistling at 3:53 – 4:45 poses a particular challenge for the balance with the tape. The live whistling should not be covered by the whistling on the tape.

Instructions on expression
In page 1 individual text syllables should be emphasised while uttering the sottovoce mumbling.
The original parlato sottovoce character of “esse comprobatur” between 8:50 – 8:53 was changed to: speaking loudly and ironically.
The in 10:50 – 11:02 should be articulated loudly and in very short staccato.

López López, José Manuel (2019): Interview. Conducted by Germán Toro Pérez, 16 January 2020. [quoted as Interview]

López López, José Manuel (no year): Note de programme. Online: http://brahms.ircam.fr/works/work/10280/ [last accessed 25 March 2020]

Recording: electric phoenix
http://www.electricphoenix.darylrunswick.net/audio