Sound design


Earth Baby is the title of an installation for a major survey of the work of Japanese artist Tomoko Konoike. The installation involves a large revolving sculptural form - a child-like head, emerging from a small mountainous crag. Like a baby's head being born from a volcanic eruption, the head sits atop the form, spinning around with mouth wide-open. The volcanic rise is propped in place by wooden poles similar to those used for rural and costal land shrines, while the floor is covered by thick swirling rope also employed in ceremonies celebrating the elements. The baby's head is covered with tiny shards of mirror, and an array of lights focused on the spinning head creates a dizzying swirl of overlaid light beams circling the space. The installation is in a 13m x 13m room, and access to experience the work is via a suspended platform running along 2 walls of the space.

The sound design for Earth Baby is composed, engineered and produced by Philip Brophy. The installation mix is in 10 channel audio, with an additional 2 channels of sub-bass. Earth Baby opened at the Tokyo Opera City Art Gallery for the exhibition Inter-Traveller curated by Shihoko Iida.


Sculptural installation - Tomoko Konoike
Music/sound composition & mix - Philip Brophy


Premiere exhibition - Tokyo Opera City Art Gallery, Tokyo


Discussions for Earth Baby commenced in December 2008 in Tokyo. Tomoko Konoike outlined the formal and physical aspects of the central sculpture and its installation. From this outline, discussions centred on three dimensions related to the installation:
1. the sound that the open-mouthed baby would make
2. the sound of the 'earth' from which it came
3. the sound of the 'new world' space into which the baby was being born

These three dimensions (which also relate to time phases of the baby's 'travel' from the centre of the earth to the surface in the gallery space) then formed the basis for developing sound textures and sound atmospheres, each of which would relate specifically or collectively to aspects of the installation. At this stage the dialogue was quite conceptual and philosophical, but it also leaned substantially on the oeuvre of Konoike's work which deals with transitional states, phantasmagorical landscapes, cataclysmic ecology and psycho-sexual imaginings of the human form. The general directives were:
1. have some sounds slightly rising in pitch in waves, so that the sound is always ‘rising up’
2. have most sounds build in waves and periodic surges of sound, like the earth is breathing
3. have sounds sourced from the Southern Hemisphere – so that they make their way to the Northern Hemisphere – like the baby’s voice rising out of the volcanic mass

A set of sounds were then decided on to be recorded which related to the ideas discussed:
A: Wolves howling
B: Cicadas from Queensland tableland (incredible waves of cicada noise as the cicadas speak to each other across a large valley or forest area)
C: Tokyo train carriage creaking (sounds like a baby crying softly in its sleep)
D: High-pitched chord of voice-like hollow ringing from train travelling fast (sounds like a group of babies cooing)
E: Rubbing/rasping of ceramic cups to effect ‘volcanic rock voice’ texture (this can be performed to sound almost like a deep voice speaking an unknown language, mostly made of vowel sounds)
F: Rolling logs along ground (to simulate earth moving)
G: Rolling stones along wooden floor (to simulate thunder)
H: Creaking rope (to relate to the baby stretching upwards from the ground)
I: Bubbling mud (to suggest the heat of the molten earth)
J: Steaming kettle (to suggest the heat of the molten earth)
K: Types of whirling strings/materials in wind – experiment with circular ‘bull-roarer’ effects (the ‘bull-roarer’ is an Aboriginal ‘sound-making’ instrument)
L: Various marbles spinning around wooden and earthenware bowls (this would sound great spinning around the room)
M: Water pouring thinly onto rubber covering over drain to create resonant drum rolling texture (this is a strong resonant sound with a rich deep tone)
N: Unevenly bottomed ceramic cup or bowl rattling on a wooden table until it comes to rest
O: Drum kit rumbling produced with fans, vibrators, etc. (sounds like the drum kit is alive and ‘playing itself’)
P: Various sheets of hard material rumbling produced with fans, vibrators, etc. (again, sounds like the material is vibrating in response to deep underground activity – like distant earthquake shaking)
Q: Gurgling of water slowly being drained from a deep bath (this makes a voice-like gurgle) R: Fireworks from festivals and celebrations S: Breathing through hollow tubes and cylinders T: vocal singing of sustained pitches to multi-track for choir effects

Not all of these sounds ended up in the final version, but the overall feel and intonation of these type of sounds and their sculpting largely became the basis of the completed piece. Mono and stereo recordings (both in the field and in the studio) were assembled, then sampled into sections and fragments to generate 'performable' sounds via MIDI modulation.

In addition, a series of recordings of real-time performances using bowed cymbals, extended guitars, and hand-played drum skins were produced. These more 'musical' elements were performed in large arcs of 'breathing rhythms' in accordance with the other sounds. Finally, a 'sample choir' was constructed from singing a 2 and half octave range of rising whole tones which were overlaid in a series of stereo layers. At any one point, the left side would be a true-pitch recording of one tone, while the right side pitch would be a pitch-altered version of the sampling of an alternative tone. The result is a slightly shimmering mix of 'real' and 'artificial' voice. A ten-note tetrachord sequence was composed in preparation for splitting the multiple voice recordings across multiple speakers in the final installation. The general blurring between 'real' and 'artificial' instrumentation and aural processing - ie. of making drums surge like waves and wave recordings rhythmically pulse like drum rolls - is an idea inspired by Toru Takemitsu's ground-breaking score for Maskai Kobayashi's Kwaidan (1964).

A 'suite' incorporating all these sounds was composed and produced in stereo across March and April in 2009 in Melbourne, then uploaded online for Konoike to check in Tokyo. Upon hearing these, Konoike decided on incorporating some ocean wave sounds. In May 2009, these new sounds were added to a new mixage in Melbourne, blending in well with the 'breathing rhythms' which informed the way in which all the sounds were performed.


In June 2009, further modifications were suggested by Konoike which resulted in a more pared-down version of the 'suite' originally submitted. Final changes were implemented in Tokyo in July in the lead-up to the final installation. In Tokyo, all stereo versions of the sounds were reconfigured into 4, 5 and 6-channel versions. The overall sequencing, timing and spatialization of the sounds was developed in the actual space, taking into account the tempo at which the baby's head rotates, and the overall scale and immersiveness of the space. Concurrently, a speaker configuration for the space was determined.

The speaker placement was crucial in allowing the 'breathing rhythms' of all the sounds to shift quite dramatically yet 'organically' throughout the space. Basically, at any one time there is about 3 to 4 sound components audible, and each of them occupies its own unique 4, 5 or 6-channel matrix within the 10-speaker spatialization. So while the overall feel of the piece is minimal in terms of orchestration and layering, the spatialization is quite active and relates directly to the giddying effect of the mirror beams continually circulating through the space. Final suggestions made by Konoike upon hearing the sounds in the space involved inserting occasional pauses, silences and periods of emptiness. The subtle sounds of light rain and distant wind were also layered into the final mixage. At this point, 2 separate DVDs of Dolby Digital 5.1 sub-mixes were produced. A DVD-synchonizer then plays the dual DVDs to generate the 10-channel composition. The finalised running time of the looped composition is 14 minutes and 24 seconds.