Landscape with figures II

Whispering room


John Cage visited an anechoic chamber and noted that in a room without sound, sounds of the body become very clear.


"There is no such thing as an empthy space or an empthy time. (...) For certain engineering purposes, it is desirable to have as silent a situation as possible. Such a room is called an anechoic chamber, its six walls made of special material, a room without echoes. I entered one at Harvard University several years ago and heard two sounds, one high and one low. When I described them to the engineer in charge, he informed me that the high one was my nervous system in operation, the low one my blood in circulation. Until I die there will be sounds. And they will continue following my death. One need not fear about the future of music."[1]


I early sketched situations which could form contrasts to the writing for full ensemble. A shadow situation could arise with whispering voices and electric tinnitus sounds. I made references to Gerard Grisey's work Jour, contre-jour (1978), [2]  opening like a high tinnitus and deeper rumbling sounds, against occasional irregular heart beats and breaths.


Voices should be whispering, slowly or rapidly, with nervousness and interruptions. The singers could whisper extremely fast, and recorded whisperings could move to even more extreme speeds, into a suggested 'buzz of voices'.[3]  Associations move towards grasshoppers already present in the sound installation fragments.







































































One idea was to take the Cage situation literally and record sounds of a body in an anechoic chamber, though I ended up creating these whispering rooms by other means.


The sketches suggest sound phenomena which are difficult to notate, in transitional zones between the whispering voices and their digital transformations. Musical notation was set out to be an important tool for the project,[4] though not necessarily the only tool. I decided to open these parts for whispering improvisations over extracts from Finnegans wake. The whispering voices can be manipulated in a multitude of ways approaching the imagined electric, nervous, bodily or artificial sounds.


Pages of the Joyce texts were recorded in studio by Song Circus,[5] in 3 versions:


  • Whispered and clearly audible text.
  • Extremely rapid whispered text.
  • The texts are used as a source for improvisation. Several parameters can be out of sync with the text:
    • Dynamic shapes.
    • Inhaled and exhaled breath gestures.
    • Hand tremolos in front of the mouth can obstruct the text.


Eva Bjerga Haugen improvising on Finnegans wake.




More than 100 fragments were made for the whispering rooms. While making these 100 versions, registers and synthesis methods were varied. These are some of the versions:


  • All whispering recordings are spectrally transposed,[6] blurred[7] or arpeggiated,[8] and generally speeded up. Through this some of the nervous qualities suggested by the sketches were achieved.


Joyce text 2.




  • All whispering recordings are either transposed or pitch shifted,[9] speeds vary from original speed to very rapid. Spectral pitch shiftings add new qualities to the sounds; components of the spectrum are not moved in proper overtone proportions like in transposition, making the sounds more artificial. Resonance frequencies and percussive attacks arise. In low registers shifted whispers sound like bouncing ping pong balls, in high register brilliant sounds appear.


Joyce text 12.




Joyce text 14.




Joyce text 16.





  • Recordings are transposed and pitch shifted, all played at rapid speeds. This comes closer to ideas in the sketches. The live singers do what is physically possible, and the manipulated version bring the whispering to impossible speeds.


Joyce text 43.




  • Whisperings are morphed[10] towards grasshopper and tam-tam sounds. The sounds are more artificial and electric sounding from the FFT distortions. This approached some of the electric nervous system sounds imagined  in the sketches.


Joyce text 61.




Joyce text 74.




Joyce text 79.




    Joyce text 80.




  • Recordings are only pitch shifted in the highest registers. After a while I found these sounds interesting.


Joyce text 146.




Making 100 versions of an idea gives a possibility to experience variations in how the processing affect sounds, and gradually guide the results towards more precise textures. This is quite similar to practising an instrument.


The fragments of processed sounds randomly are triggered in a sound installation situation. Categories of fragments were:


  • The new Joyce text fragments.
  • Existing fragment types from the Landscape with figures I sound installation:
    • Breaking.
    • Distorted breaking.
    • Ammons water.


The score suggested degrees of clarity, distortion or improvisation of the text. Song Circus chose to distribute ideas differently between the singers. Pages from Finnegans wake were held against the improvisation scores.




































Detail of photo by Henrik Beck.


AE Whispering room 9

The live whispering voices went through additional live processing through MaxMsp:


  • 5 voices are routed to 5 different time variable delays, at random walks varying between 1.2 and 27 seconds.
  • A FFT amplitude filter lets through the loudest frequencies . Filtering vary by random walks. Only sounds with high energy are passed. Synthesis parameters are randomly changed at random time intervals. This varies degree of distortion of the inputs. The result passes to final delays and moves through further processes.
  • The filtered sound goes through an FFT feedback process where settings are in random walks. This added more electric and metallic sound qualities.
  • Results of the filtering and feedback was supposed to go through an additional variable high frequency FFT pitch shifting.[11] The result could well create the imagined tinnitus sounds of the initial sketches. I had to skip this last step, as I was reaching a CPU limit through all processing, sound playback and spatialization involved. The highest priority should be stability throughout the performance.
  • The results are routed through 6 more time variable delays, at random walks between 1.2 and 27 seconds.


Routings through these processings are done by matrixes, in different patterns for each Whispering room fragment. Variable routings, variable delays, filtering with variable degree of response and sine wave distortion, all contribute to unpredictable and continous high frequency textures from the whispering singers.


The whispering room sketches are principles rather than fixed materials, through improvisations and unpredictable processing. Results can approach the initial ideas or produce something new. 






































[1] Cage, 1973, Silence, Lectures and writings by John Cage, p. 8.

[3] Ruben Sverre Gjertsen, 2011, Between instrument and everyday sound,p. 5.

[4] Ruben Sverre Gjertsen, 2011, Between instrument and everyday sound, p. 4.

[10]  The Csound opcodes pvcross, pvsmix and pvsfilter.

[11] The MaxMsp gizmo~ object.