"Soto l'imperio del posente prince" is a composition by Jacopo da Bologna, and a rapidly ornamented version is part of the Codex Faenza manuscript. Rather than composing a new piece based on this, we used the original score from Codex Faenza, the Jacopo da Bologna and the Codex Faenza version together, while searching for a vocabulary of timbral abstraction. I brought a list of four situations to rehearse.

    •    O: "Ordinario", close to idioms often used by early music ensembles.
    •    I: Fidels: Light bowing at the bridge, flautato, counterbalance the weigt of the bow. (Common technique in works by Helmut Lachanmann. ) Alternate with bowing on the body of the instruments. Recorder: Pure light noise sounds. Voices: Whisper only, inhaled/exhaled. (Joyce texts.) Organetto: Very fragile noises or barely audible flageolets. All: Find timbral micro articulations and differences within the pure noise range.
    •    II: Fidels: Light circular bow, shimmering flautato sul pont. Use natural harmonics or contantly vary microtonal pitches. Recorder: Prepared fragile buzzing sounds. Voices: Half voiced to whisper. Multiphonics and other vocal noises. (Joyce texts). Organetto: Glissando, bending, fragile sounds.
    •    III: Fidels: Rapid brusing sounds, spiccato arpeggios up and down on varying harmonics or slowly evolving glissandi. Brillante from ppp to mf. Recorder: Various flutter types on obscure rapid phrases, transitions between noise and tone, possible on/off preparations. Fantastico. Organetto: Contrast to the other instruments, instead of rich surfaces, choose slowly evolving ideas, sustain points, "drones". Voices: Short sudden outbursts, bubbles of sound, create silences and avoid establishing patterns.

Lets add some more comments about these timbral situations.

  • The ordinario sound is the common early music sounds, clean non vibrato tones without any extended techniques. We called this the 'surface', while the noise sounds were 'submerged'.


  • The first abstracted version has a pure noise or whistle sound. It can be a challenge to avoid pitched sounds on the instrument. Helmut Lachenmann suggests bowing lightly at the bridge or at the body of a string instrument. 1


  • The second abstracted version was a shimmering, with emphasis on overtones, buzzings and instabilities.


  • I called the third abstraction the "Paganini" version, or related to the way Sciarrino has used abstracted virtuosity from the past in his 6 Caprices for solo violin. 2 He specifies light vertical brushing with the bow. I will refer to Sciarrino on this, while it is not common for composers to quote all their influences. It is more comfortable to present the loots as your own. Any stolen sound need to be integrated in a new context, it is not just about inventing new sounds, but establishing a musical language.

My timbral suggestions were schematic, as an exercise to find sound palettes for the ensemble. When those were established, they could be used fluently in transitions, patterns of 'surfacing', 'submerging' and erasing. After initial suggestions from me, detailed decisions were quickly taken over by the ensemble, and what you will hear is very much their versions of the piece.

In the Plamenac edition, measures 37-39 of the diminution line were marked as inedible in the manuscript. 3 I suggested using this opportunity to perform erased sound as white noise.

Jostein Gundersen put together a complete version of the diminutions using a newer facsimile edition. Through the use of infrared light, the missing measures were indeed readable. 4

Our first performance of this piece was linear, we were performing an existing score with altered sound qualities. This approach to quotation, a washed out and erased reading, has been used by Gérard Pesson in his "Nebenstück" for clarinet and string quartet. In this filtering of the Ballade op. 10 no. 4 by Brahms, Pesson uses the whole piece in chronology, erasing notes, replacing notes with noise and alternative sound qualities, and streching a few points of the piece. 5 It is linear and true to the original. For the "Wheels within wheels" project, it was useful to go closer to an original medieval composition after the  extensive distortions and non linear readings of "Nuper rosarum flores".

When we were writing the project application for the "Wheels within wheels" project, Morten Eide Pedersen suggested to use Score Following for interaction between musicians and live electro acoustic treatments. Sadly, he did not live to be part of the project. Antescofo from Ircam is today a highly developed language for score following, which listens to a performer, follows a score, and sends synced messages for any triggering and treatment of sound. When used well, the computer should be able to act like a co-musician and follow tempo variations.

I studied the Antescofo language and implemented in it my OM-Ruben library. The demo patches concerning Antescofo uses musical notation for followed and triggered notes, attaches messages for various effects and envelope shapes, and exports code understood by Antescofo.

Lets take an example of what this looks like.

 ;Exported score and actions from Open Music for Antescofo

;Detection settings (uncomment to use it):
     antescofo-mess gamma -1.0, tune 465.0, pedal 0 ;, nofharm 10, suivi 1 @local
BPM 96
NOTE 0 0.16320002  

NOTE 6698 3.4688  

   GROUP Phrase1
             sampler 7400 977 121 4
          1.467 sampler 7400 277 114 4 ;Comment or parameters
          0.41400003 sampler 7604 291 107 3
          0.43799996 sampler 7400 310 100 2
          0.46350002 sampler 7192 313 92 1

NOTE 6698 2.024  
   KILL Phrase1
   GROUP Phrase2
             sampler 7604 210 85 4
          0.315 sampler 7400 211 78 2 ;Comment or parameters
          0.315 sampler 7192 212 72 2
          0.31800002 sampler 7400 319 68 3 ;Comment or parameters
          0.47849995 sampler 7604 314 65 2

NOTE 6490 1.9104  
   KILL Phrase2
   GROUP Phrase3
             sampler 7690 306 61 1
          0.459 sampler 7898 301 57 2 ;Comment or parameters
          0.45 sampler 7690 295 54 3
          0.44250005 sampler 7604 292 50 4
        curve @action:= morph1 $x    @grain:=0.001s
            {0.493} @type "sine_out"
          0.5999999  {0.41833332} @type "sine_in_out"
          0.60000015  {0.6586667} @type "sine_in"
          0.5999999  {0.492} @type "sine_in"
          0.5999999  {0.23411112} @type "sine_out"
          0.5999999  {0.47055557} @type "sine_in_out"
          0.6000004  {0.77766675} @type "sine_out"
          0.5999999  {0.5676667} @type "sine_in_out"
          0.5999999  {0.65888894} @type "sine_in"

The NOTE events are tones the Antescofo Max object should expect from the microphone. When it is detected, the following messages are send to be used by other parts of the Max patch. This example is a script for "Le Ior" from Codex Faenza. The followed part is the tenor line, and the computer will attempt perform the diminution line with the live musician. The 'Phrase' groups send note information to a sampler, with all notes happening over the current tenor note. If the musician is playing too quickly, the computer should KILL the past phrase and move over the the next one. At the last NOTE in our example, there is also a curve named 'morph1'. It can be used for morphing parameters like suggested here, or any processing of the input musicians, for instance filtering, ring modulation frequency, or spatialization shapes.

With some latency issues, it could have worked more elegantly the other way around; the live musician plays a series of notes and the Score Follower cues in effects at particular notes. The latter is how it is mostly used by Pierre Boulez his the piece "Anthèmes 2" for violin and electronics.

I prepared "Soto l'imperio del posente prince" for this type of score following. One of the musicians would have to perform the tenor line clearly enough for the Score Follower to understand it and add the diminution line. Curves were controlling morphing and filtering. Ring modulations added abstract glitches to sampled celesta, viola pizzicato and marimba.

For those interested in details, the live processings were morphings from Spectral Toolbox, iana~ from Ircams Max Sound Box. Variable delays were added to avoid a direct reaction to the live performances, a strategy used in many live electronics works by Luigi Nono. Sounds were spatialized with the DBAP algorhithm 6 through Ircam Spat, distributed over a circle of 8 speakers. Some parameters were automatized through Score Following while others were controlled live during the performance.

On the 2th of may 2017, Gallery 3,14, Bergen, the "Wheels within wheels" research group arranged a workshop with Gerhard Stäbler as an invited guest speaker. Ensemble Currentes with Bodil Rørtveit performed "Soto l'imperio". The audience was encouraged to move around, interact and make suggestions for the performance. One suggestion was not just doing a linear performance of the piece.

Improvising "Soto l'imperio del posente prince"

Ensemble Currentes received friction from multiple delayed and processed versions of themselves performing the piece, including two guest singers not present in the final concert. The delays were not in sync with the tempo like for a regular canon, for the effect of smearing the piece out in time. This was surprising for the ensemble, and possibly obstructing what they had planned to do. Abstracted versions accumulated new sediments, the listener will face an archeology of the piece. Jacopo da Bologna and Codex Faenza can be found among the bottom sediments. Robert Smithson used the terrain in a direct way in is 'earth art'. We can borrow his descriptions as metaphors.

"The earth's surface and figments of the mind have a way of disintegrating into discrete regions of art. Various agents, both fictional and real, somehow trade places with each other - one cannot avoid muddy thinking when it comes to earth projects, or what I will call "abstract geology" ". 7

The stretched version.

We can see Kjetil Møster obstructing the sound of Josteins Gundersen's recorder, and Ruben Sverre Gjertsen running the sound. I recorded the individual tracks of this performance, and used them for the mix we can hear here, and further, more extensive processings for the final "Mouvance III: Distortion" concert.

During the Artistic Research Forum in Fredrikstad, 18/10-20/10 2017, we made a new interpretation of "Soto l'imperio". Following the suggestions from the workshop in Bergen, we performed a stretched version rather than the whole score as written. Jostein Gundersen, Hans Knut Sveen and guest singer Berit Norbakken Solset decided to perform from the beginning, and augment note values of a few measures. These notes were filled by oscillations, breath sounds, sub harmonic voice, timbre trills and other abstractions. This stretching opened a new dimension to this piece.

There was no linear tenor part to follow. An Audio to MIDI analysis was turned on and off at will, as a replacement for the first score following setup. I kept individual sound tracks for "Mouvance III: Distortion", where we also performed a stretched version.

The first performances involved live electronics setups and Score Following. I decided to use neither of them in the final concert, just a few variable filters and delays to blend the live performance into preprocessed sound tracks. I met limitations with live electronics, and was ready to use more heavy non-real time processings with Csound and high order ambisonics. I will give descriptions of these during the discussion of the sound installation elements.

We had two versions of "Soto l'imperio del posente prince", for the poor potent prince, at our final concert.

The linear performance.