Content uploaded by Roland Cahen
Author content
All content in this area was uploaded by Roland Cahen on Jun 21, 2022
Content may be subject to copyright.
Kinetic Design
From Sound Spatialisation to Kinetic Music
Roland Cahen
Centre de Recherche en Design (CRD)
Ensci les Ateliers – École Normale Supérieure Paris-Saclay
roland.cahen@ensci.com
Abstract.
This paper explores the process of kinetic music design. The first part of this
paper presents the concept of kinetic music. The second part presents the sound
design and compositional process of this type of music. The third part presents
some excerpts from the composition logbook of a piece called Kinetic Design to
illustrate the process of kinetic design as work in progress.
This paper focuses on the question of sound spatialisation from a theoretical,
as well as an empirical, point of view, through the experience and experiments
of an electroacoustic music composer trying to make the imaginary concept of
kinetic music real. It is a form of research by design, or research by doing. The
kinetic design project examined here is the first time an experimental approach
of research by design has been applied to kinetic music
Keywords: sound spatialisation, kinetic design, kinetic music, electroacoustic
music, sound design, design process, composition
1 Introduction.
Kinetic music aims to produce sound choreography where the sound is diffused. Hence,
it uses sound spatialisation in such a way that both the composer and the listener focus
on kinetic aspects of sound, in opposition to using spatialisation only illustratively, or
for rendering effects. In kinetic music, like theatre, dance or visual arts, each zone,
position or direction can take on a musical value, a form of density that the sound space
itself embodies. Kinetic music wishes to add a new form of expression and composi-
tional methods to existing spatial music concepts and techniques. Sound spatialisation
has already been the subject of abundant literature, the focus of this paper is to demon-
strate the specificity of kinetic music. Of the literature on spatial sound, much has been
written on the subject, such as generalities (principles, philosophy of space and music,
history) numerous tools and techniques, analyses of musical intentions and abstractions
about spatial figures, but very little has concentrated on the auditory experience, spatial
sound aesthetics and none on the design process.
2
(This article explores the hypothesis of kinetic music as a new process of composition
and analyses its compositional process using my piece, Kinetic Design. The term design
is used hereafter in relation to the creation process. The creative process explored here
is ‘research by design’ or research by doing [6]. This form of research has a long legacy
and is not novel in music. Indeed, Pierre Schaeffer developed his experimental ap-
proach by ‘doing and listening’
1
. However, the kinetic design project examined in this
paper is the first application of an experimental approach to kinetic music. By exploring
the different motivations that drove the project, the successful and less successful ex-
periments, and looking at extracts from my composition logbook, this paper tries to
shed some light on some of the basic concepts and methods for kinetic sound design
and composition. Kinetic Design was commissioned by INA-GRM, composed in octo-
phony and performed on the acousmonium for the first time on January 20th 2019 at
the MPAA Saint Germain (Paris).
2 The Characteristics of Kinetic Music
Kinetic Music aims to shape empty space with a choreography of sounds. However, a
body of sounds would obviously be a kind of simulacrum
2
, quite different from the
human body or existing material objects. Composing and listening to kinetic music
means focusing attention on spatial differences and similarities, as essential parts of
kinetic musicality. Kinetic music adds new values to orchestration: spatial plasticity,
incarnation and corporality, bringing an orchestralisation
3
of electroacoustic sounds.
Kinetic music could be a breakthrough for electroacoustic/acousmatic music, anticipat-
ing new formal experiences and enabling new musical styles to emerge. However, to
achieve these beautiful promises, the audibility of kinetic effects must be guaranteed.
2.1 Spatial Sound and Music Existing Work
There is a vast literature about sound spatialisation in the domain of music technology,
electroacoustic music and perception. New rendering techniques procure nowadays a
better sensation of sound incarnation thanks to new techniques such as WFS, High Or-
der Ambisonics, dynamic pan and routing and room simulation. Simultaneously, spati-
alisation tools, editors such as the Ircam Spat, Panoramix, ICST Ambisonics, MaxMSP
mc., and animation tools such Music Space, Acousmodules, Iosono Animix, GRM
Tools (Spaces) and Iannix give most multichannel DAWs (digital audio workstations)
and 3d real-time game editors (such as Unity3d) facilities for editing and automating
sound source positions and motion. At the same time more international scientific, ar-
tistic and audio production centres are working on large multichannel sound devices to
experiment with lines, arrays, matrix, domes and other sets of speakers. The existing
1
Musical research approach by Schaeffer [11]
2
Simulacrum refers to representation in ancient Greek literature and philosophy. This term was
also used in XX century by Schaeffer [12]
3
A neologism meaning making something orchestral which was not originally.
3
literature about sound spatialisation describes perception, concepts, descriptors and
techniques.
4
Object based formats allow to give sound sources 3D(xyz) theoretical po-
sitions, that can be rendered to any standard multichannel soundfile format and in any
real space, independently of the number and positions of the speakers. The SSMN Spa-
tialization Symbolic Music Notation is an abstract and geometry toolkit for writing spa-
tialised sound movements on instrumental scores. Spatial interactions with physical
gestures are also explored at CIRMMT
5
and at Ircam
6
.
2.2 Audibility as a Premise
Most spatial music rests on so-called trajectories, easily visible on sound editors’ inter-
faces, but rarely audible in situ. However, spatial shapes, i.e. the compositional building
block of kinetic music, should be clearly audible in order to be operational. Even if
music can be heard when spatial sound shapes are not obvious, it is no longer kinetic
music. What we could call kineticality, as a measurement of kinetic audibility, could be
defined by how many audible elements of kinetic sound content or qualia
7
are lost in
the spatial reduction process
8
. Unfortunately, musical audibility is more difficult to
measure than individual sound-effect audibility. Spatial sound and music audibility can
change a lot from one listening context to another and from one listener to another, for
example because of the hotspot effect,
9
which some techniques, such as real source
positioning
10
and linear panning
11
can reduce.
12
4
Blauert, Bregmann, Rumsey, Jot, Warusfel, Pachet/Delerue, Cadoz/Luciani, Brümmer, Du-
chenne, Pottier, Vandegorne, Schumacher, Schacher, Baalman, Orlarey etc.
5
Schumacher with OMPrisma lib for open music
6
Bevilaqua, Schnell, Lambert CoSiMa project.
7
See the kinetic music qualia below.
8
The reduction process consists of downmixing or converting multichannel sounds to stereo or
mono.
9
Only listeners placed in the hotspot position can hear the spatial effects properly, e.g. equidistant
from the surrounding speakers.
10
Real source positioning, as opposed to virtual positioning, techniques consist in placing each
sound element on only one track / speaker. It has been used since the very beginning of spa-
tialisation (multiple mono) and considered theoretically as a spatial compositional approach
by various artists such as Pierre Boeswillwald in Octophonie Delta P (1991-94) with the con-
cept of “octuor” or Benjamin Thigpen [13].
11
Linear panning stands for panning between two next in line speakers, works also for a circle
where the azimuth controls the position.
12
The necessity to listen to spatialised sounds from a small point in the center of the device. The
hotspot effect is more important when using virtual source positions than with real source
positions.
4
Table 1. Easily Audible Kinetic Sounds/elements vs. Hardly Kinetic Ones
Audible
Inaudible (or less audible)
Recorded/synthesised im-
pulses/attacks
Sine waves or tonic sounds
with poor spectral complexity
Various noise dynamic mod-
ulations (filtering, granular-
ity…)
Soft attacks are less precisely
positioned
Tonal sound with rich com-
plexity
Spatial complexity blurs ki-
netic sense
Reduced spatial complexity
up to 2/3 voices
Complex movements
2.3 Kinetic Design Qualia
Here is a short list of kinetic qualities experienced in Kinetic Design. It is difficult to
separate concepts, sensations or experiences from sound shapes.
• Immersion, position distribution, sound mobility.
• Punctual sounds and impulses: position (referent / changing), position blur, motion
by re-iterations, motion by elementary transition, referent positioning, accentuation.
• Sustained sounds: motion by transitions (trajectory) between positions, spatial ac-
centuation.
• Movements: pointed, traces drawn through space, rhythm, oscillation, swing, bounc-
ing, hold and release, etc.
• Mass: extension / diversity, deployment/folding, spatial resolution.
• Plasticity, incarnation, sound corporeality.
• Explicit visual reference (visualisation) vs. implicit or metaphorical (materialisation,
kinaesthetic shape, imagination).
In order to better understand where this typology comes from and how it has been es-
tablished, section 3 will present some of the concepts of kinetic design qualia through
the process of experimentation and composition, referring to the composition logbook
of Kinetic Design.
2.4 Visual Imagery
Designing geometrical patterns of sounds, such as a sound trajectory, is not sufficient
to produce kinetic sensation, but our imagination seems nevertheless to shape our rep-
resentation of kinetic sounds. On the one hand, imagining these spatial sound shapes,
our natural inclination is to project a visual grid on the auditory perception and to try to
draw sound in space. However, by doing this, we might risk distracting the listener from
the auditory experience. The act of hearing a sound should be sufficient in itself, and it
is necessary to keep it at a distance from visual design, or tools for visualisation, so as
to privilege audibility, i.e. auditory experience and sensation. On the other hand, “ki-
netic sound design” refers to visual imagery. Some visual metaphors are realistic and
5
obvious, such as the big swing, or the little soldiers (Table 2) marching around the
listener], others are more abstract, such as diversity variations or positional contrapuntal
impulses (Table 2 Dots & Lines). This is why kinetic sound design belongs to the dis-
cipline of design, since sound shapes emerge from the sound itself referring to our spa-
tial representation of the world, which is mostly visual.
In this article the term design is also used in relation to the creation process which this
article presents through a “research by design’ approach.
3 Composition Device, Design Process and Workflow
A description of a kinetic design and compositional process follows. This enquiry en-
compasses theoretical and then applied research. Since section 4 contains a presentation
of my experience and methods of kinetic design composition, as a concrete illustration
of this process, an outline of the composition follows in order to help the reader situate
the different sections of Kinetic Design discussed below. The workflow depicted below
is only one among others, but it focuses on i) kinetic audibility and ii) spatial composi-
tion and spatial authoring possibilities at every stage of the compositional process from
its conception to performance.
Fig. 1. This schematic offers an overview on the design / compositional process. The concep-
tion phase consists of imagining audio kinetic ideas and making sound sketches or Max models
for experimenting with them. Only elements with kinetic musicality and audibility are kept for
composition. Otherwise ideas are improved or abandoned. This evaluation, which also happens
at further stages, is the condition of existence of kinetic music. Formalisation is the proper com-
positional part of the project, where sound elements are performed, recorded and organized into
a music composition. During this part of the process, it is essential to facilitate modifying the
sound elements.
3.1 Ideation
Planning the kinetics from the very beginning of a composition project enables us to
put movement at the heart of a project, whether the kinetic plan concerns sound design
or the whole composition. Adding spatialisation once the whole composition is
6
finished, or even once sound elements are recorded, is generally too late to achieve
kinetic creation consistency. Design ideation techniques exist and flourish in design
thinking literature. Some can partly be applied to art creation as well as functional de-
sign. Here are two complementary methods which can be used to sonagine
13
kinetic
sound objects:
i) An inductive method which starts from an imaginary metaphor. For example,
“Swings” (see Table 2.) inspired by Edgar Allan Poe's the pit and the pendulum, imag-
ining a giant pendulum, whistling by passing near with trajectory. Here the idea is typ-
ically physical or visual.
ii) An experimental derivational process which imagines an abstract tool for moving
any sound source materials. By experimenting with various sound sources and perfus-
ing the tool, it is possible to produce lots of musical materials, out of which the most
interesting can be kept for creating musical sequences.
Ideas can sometimes reveal themselves as being well sonaginated, but also other times
less so because they are totally abstract and unrealistic, too visually oriented or inap-
propriate. In reality these two methods are often used successively or combined.
3.2 Experimentation and Sound Production in Max MSP Custom
Octophonic Tools
How can a sound idea like “Swings” (Table 2) be rendered? How about synchronising
between variations of timbre, position, spectral mass and spatial spread/diversity?
DAWs environments are quite rigid, inconvenient for creating sounds, especially with
dynamic parameters’ variation when including spatialisation. Circumventing the com-
plexity and haziness of channel management can easily be time consuming. Therefore,
it is more efficient to develop fast MaxMSP sketches for each idea. These procedural
models can also be improved during the experimental phase until they reach a sufficient
precision to fulfil functional and expressive requests.
Two examples are described here:
- OctoLine allows to control a set of parameter’s variations for a single stream,
using break point functions to modify dynamically the following: volume, po-
sition, a filter cut-off frequency, spread and spatial diversity.
14
Octoline there-
fore produces variations in spatial extension. It uses mostly noise or simple
generators the instances of which can be easily differentiated.
- KDvector launches polyphonic spatialised sound vectors triggered by MIDi
notes. Time dynamic parameters or parametric vectors can be set before
launching them, such as sound type volume curve, position curve, filter cut-
13
Sonagination is a neologism invented by sound designers to draw a distinction between visual
imagination and auditory imagination.
14
Spatial diversity is the desynchronization, or the originality, of the different voices. It has been
used by Charles Verron [14] in his PhD thesis to simulate natural immersive sounds such as
rain. Diversity equals 0 when 8 sound voices are mixed together and played on one or more
speakers according to the spread value. Diversity equals 1 when a different sound voice plays
on each speaker. Between 0 and 1, positions spacing of 8 the sound voices vary continuously.
7
off frequency, filter quality and spread. Diversity is ensured by multiplying
vectors units.
3.3 Recording Sound Elements
During the process of developing MAXMSP patches and producing kinetic shapes, se-
quences of octophonic sound are recorded. Each sequence is performed live or autom-
atized to produce the desired kinetic sound. All elements are recorded in multichannel
with their spatialisation embedded and will be manipulated later on in multichannel.
3.4 Kineticality Evaluation of Sound Elements
Before going further in the composition kineticality
15
is evaluated to validate sounds
the kinetic expressivity of which are audible and thus to exclude those that are not
16
. A
proper evaluation would require a large listener panel, but a simple listening evaluation
is better than nothing. Therefore, waiting a few days before re-listening helps to step
back and allows us to forget the original perception. François Bayle use to say “il faut
faire fonctionner l’oubli “
17
, in order to recover primary unbiased musical sensation.
Such a subjective evaluation would not satisfy scientific evaluation criteria
18
but seems
good enough for creating music for which the main evaluation is the final performance.
3.5 Editing and Assembling Pre-recorded Octophonic Elements in a
Multichannel DAW
19
to Compose Musical Sequences.
Once sound elements are recorded and their kineticality validated, a selection of se-
quences or parts of them are imported in the DAW. This is where the main composition
is achieved. It is important that DAWs allow the following features:
• Accept mono, stereo and multichannel audiofiles
• Edition of multichannel tracks: splitting, editing and merging. This essential feature,
which allows to modify easily and precisely previously spatialised sound elements,
does not exist as such in most DAWs.
─ Splitting for edition: easy switch from grouped tracks to individual tracks
─ Easy edition of the individual tracks
─ Merging edited elements: easy backwards switch from individual tracks to
grouped tracks after edition
• Multichannel effects and in/out routing for multichannel effects
• In place offline multichannel effects
15
See Kinetic audibility §2.2, above.
16
Nevertheless, all the sounds of a kinetic music are not necessarily kinetic.
17
‘One should use the function of forgeting’ (my translation).
18
It should be tested by non-expert listeners using proper perceptive evaluation methods.
19
DAW is an abbreviation of Digital Audio Workstation.
8
In Avid Pro Tools HD for example, it is easy to move multitrack clips to n mono tracks
and to edit individual tracks, but in order to bring back the elements to a multitrack
clip after being edited, the group must be consolidated and then moved back to its
original multitrack.
Fig. 2. Example of Kinetic Design Pro Tools HD Cession with kinetic multitrack elements (oc-
tophonic tracks) and mono elements in mono tracks (on top tracks)
In other professional environments, such as Steinberg Nuendo, it is possible to route
any kind of track to a High Order Ambisonic bus and monitor the rendering on any
number of channel outputs. But a splitting and merging process is not available unless
whole tracks are duplicated for splitting again when merging.
3.6 Performance: on Groups of n Speakers and Effects on an Ensemble
of Speakers (Acousmonium, Sound Dome or any Other Concert
Diffusion Device)
In order to obtain the best possible kinetic perception in diffusion for all the audience,
a few criteria should be taken into account:
─ The acoustic of the venue should be non-resonant: muffled venues, studios or open
air are preferable to empty resonant rooms or churches.
─ Speakers are organised by groups of n, corresponding to the number of channels in
the music or its master soundfile e.g. 8 channels => one or several groups of 8 speak-
ers.
9
─ Listeners are not too near the next speaker. A distance of a few meters allows to
reduce the blinding effect
20
Diffusion Setup
The octophonic (8.0) octal diffusion format
21
, which is an interesting work format be-
cause it is efficient and accessible as lots of multichannel devices work by 8, can be
used. Track numbering from back left to back right allows i) to avoid an automation
break in front of the listener ii) to switch and transit easily from a circular to a linear
front screen distribution
22
.
Fig. 3. Octal or circular 8.0 distribution allowing: 1- a geographic allocation corresponding to
spatial intuition 2 – the possibility to switch easily from circular to linear diffusion 3 -rejects
automation hedges values in the back (instead of in the front in standard formats) 4 – a simple
compatibility with 7.1 SDDS standard format: L, Lc, C, Rc, R, Ls, Rs, LFE.
20
Blinding effect in electroacoustic concerts happens when as a listener you mostly hear the
speaker next to you.
21
8.0 is a common format but not a market standard, but it is compatible with 7.1 SDDS.
22
Some parts of the music composition would avoid 1/8 transitions in order to be performed on
a linear speaker array instead of a circular one. During performance, it is possible to avoid
jumps when swapping from a circular to line, or line to circular, diffusion device by choosing
the nearest speakers to ensure continuity e.g. 4-5. (Fig. 6.).
10
Fig. 4. Circular-linear transitions in performance
4 Excerpts of Kinetic Design Composition Logbook
5 Table 2. Kinetic Design is composed of 12 chained
movements. (This table is presented here to simplify
references)
Time
Title
Duration
1
0:09
Stretching - Extension
1 ' 45"
2
1:54
Swings
2 ' 21"
3
4:15
Dots & Lines
3 ' 10"
4
7:25
Pulsed
2 ' 54 "
5
10:19
Untied - Incises
2 ' 35 "
6
12:54
March
55 "
7
13:49
OtoGliss_1
1 ' 26 "
8
15:15
Little Soldiers
4 ' 19 "
9
19:34
Wave breathing
1 ' 06 "
10
20:40
OtoGliss_2
1 ' 47 "
11
22:27
Laché
2 ' 14 "
12
24:41
Final Balancé
1 ' 43 "
Thursday 19th July 2018
I am in the process of finishing the first part of my research to create Kinetic Design.
My original intention was to produce a composition, or “ballet” made up of “lines” and
“dots”:
- Lines between speakers: elementary trajectories, vectors or curves.
- Dots or punctual elements on each one of the speakers themselves, ei-
ther to start and end the musical lines, to create strong accents, im-
pulses in the music, points of support or clicking effects.
Although the idea works, the result does not fit with what I imagined (or more accu-
rately sonagined). To maximise spatial precision, I choose to work with two categories
of sounds: impulses and sustained noise. I will refer to these figures as impulse and
sustained noise spatial vector
23
(or NoiseVect).
My NoiseVect lack musicality, quickly becoming boring to the ear, almost like a dancer
stretching her arms back and forth. Worse, they do not validate the idea of a line, seg-
ment or trace, especially when they are slow. Faster motion and shorter swishes, such
as mechanical pistons or whip sounds give a better idea of a directional movement. But
I could not find a way of making them sound like lines starting from a position and
23
Vectors are elementary directional trajectories going from one speaker to another.
11
finishing on another. And if we can hear whipping sounds crossing through space, the
sounds also carry a dramatic and ridiculous musical connotation. It even seems that
line/trace impression are inversely proportional to spatial precision. In other words, the
more accurate a position, the more noise the sound must carry. However, noise is per-
ceived as large and blurred and not as a line. On the contrary, sustained sounds, reduced
in mass (FN), such as pure tones, which better suggest lines, can hardly be heard at a
precise position in space. This leads me to having to choose between Scylla and Cha-
rybdis.
Fig. 5. Impression of a trace thinness being inversely proportional to spatial resolution.
Impulses placed at a precise position on one single speaker (track) are situated well in
their position, but neither give the idea of points. They sound rather like events marked
in time and space. When an impulse is attached to a NoiseVect, at the begining or end,
it sometimes appears grouped together, sometimes not
24
. But even when impulses and
NoiseVect merge together in time and timbre, impulses are not perceived as being be-
ginning and ending positions of a NoiseVect, since the positions do not seem to merge.
Spatial perception thus seems to keep separate from the sound stream. Further research
could be done on merging spatial and sound content perceptions. Fortunately, as soon
as rhythm and musicality inhabit these sounds, I enter as a listener into an astonishing
and new sound universe. Like a sorcerer’s apprentice, I feel incapable of understanding
exactly what I manipulate. I can hear some effects, artefacts, bursts, spreads, folding
and unfolding, but I cannot manage any of these elements precisely. While writing this,
I am reminded of Pierre Schaeffer, whose journal de la musique concrete [10] has been
an inspiration for me.
Monday 27th August
Capping/unplugging effect: difficult to achieve with simple clicks, try breaking an in-
fra-bass hold sound with a strong distributed click. Try curtain effects from mass of
scattered grains passing from one side to the other opening and closing a background.
─ MAKE SILENCE, rarefy, nuance, intensify differences.
Mobility takes its value only in contrast with stationary moments. Look for motion
variation control.
24
As established by Albert Bregman’s grouping principles [4].
12
─ For the swings: hold the balance in a position creating tension, then release. Try
Ping-Pong rhythms on the origin and the destination of the pendulum (Table 2).
Fig. 6. Transitions (Left) Rear swing: transition from 8 to 1 happens in the middle of the move-
ment - (Right) Two identical iterative sounds in opposite direction, starting from the rear, cross-
ing forward and ending at the back gives this amusing picture (1 to 8 + 8 to 1)
Note that the intersection of the two iterative sound movements is not in the centre but
on channel 4. If the crossing happened in the middle, the sound would be played on
speakers 4 and 5 with a bad spatial focus and would reduce the spatial resolution to
only 7 points. With 8 channels this kind of asymmetry is inevitable, but on the other
hand, odd numbers of channels would bring other asymmetries, e.g. rhythmical ones.
Twelve channels would be a very good working module but unfortunately not standard
and more expensive.
Tuesday 28th August
The GRM studio-A is comfortable and beautifully equipped
25
. The acoustics are excel-
lent, the sound awesome, the computer powerful. A shelter from the outside world in a
hidden basement, where there are no distractions. Perfect!
I have tried many musical concepts, recording hours of octophonic sequences using
different models: plots, whips, swings, spatial loops, variation of position, diversity,
polyphony and spatial amplitude, with various sound materials: dynamically filtered
white noise, various wave generators and samples. On the whole there are some re-
markable figures, but I am struggling. Once these sound movements are produced, it is
difficult to assemble them. I tend to regress, replaying musical approaches of the late
70’s called “séquence jeu”, where the same set of gestural sounds are repeated with
pattern variations, quickly becoming monotonous. Trying other formal approaches, I
combine different parts of the composition as if they were sediments, layers of a land-
scape, as if each one of the parts of the work were stacked on top of one another. I play
with transitions, creating geological faults in the landscape, planes and protrusions,
leaving glimpses into the future or flashbacks into the past.
25
The studio is equipped with 8 Genelec 8250 monitors placed around the central working posi-
tion.
13
I realize that my sampled loops are spatially shifting because of a scheduling drift be-
tween the signal and data, which I then correct. At last, each sound element in a sample
or loop now plays in sync and always comes through on the right speaker allowing me
to make kinetic rhythms. I can also precisely control the position of each component.
I have created my own instruments with Max, which gives me a lot of possibilities, but
the difficulty with Max is that I develop and add new features all the time, and the
patches are never stabilized. Each time I try to add a new feature to the patch, new bugs
appear, and I need to debug from time to time. Moreover, it is impossible to memorize
the best presets because after each change the settings previously memorized become
obsolete. Consequently, I record audio at each step and use Max “pattr” presets only as
setting-up facilities.
Kinetic masking effect and kinetic pattern segregation - the masking effect
26
acts at the
spatial level. It is possible to play on the contrast of high-pitched moving sounds mixed
with sustained bass sounds, or on close planes in relation to more distant planes. But as
soon as two simultaneous movements or positions are exceeded, independent move-
ments become quite difficult to hear, except with very different or complementary or
eccentric sounds. Hence the use of very eccentric registers, e.g. the infra-bass only
slightly disturb the spatial perception of other sounds. Maximising temporal, spectral,
dynamic and of course spatial differences helps to separate the patterns.
Playing with the diversity parameter - diversified sound masses, i.e. composed of 8
independent sound sources, but of the same nature (e. g. 8 different white noises on
each track versus the same white noise on the 8 tracks), allow me to stabilize the move-
ments of sounds and give our ears a rest. They create a sensation of breadth and rich-
ness, like open-air soundscapes.
MAKE CLEARER - spatial complexity and diversity must be undressed, flattened or
lightened, the movements should stop. More static and monophonic moments would
reinforce the spatialised ones. Should I do this in the composition or during the perfor-
mance? Both probably. Unfortunately, the possibilities to reduce complexity without
losing content during the performance are limited.
Thursday 27th December
I am at the end of the project. The composition is more Baroque and more dramatic
than originally expected. Many bass and infra-bass sounds, clinging clicks and striden-
cies provoking inner ear interferences. Treble strident sounds are heard as they saturate
the ear itself. Strange feeling, It seems as if the sound was produced by the ear itself,
often mixed up with provoked otoacoustic emission, I do not think they are the same
phenomenon. They happen when treble sine which are distributed in frequencies and
space and are modulated slowly, typically between 1 and 4 kHz. At the end of listening,
the ear is physically tired, yet the listener has travelled to unusual places and sensations.
I believe I have succeeded in my challenge to create kinetic music, where spatialisation
plays a central role, constitutive of the musicality. There is a lot of imagination and
variety in kinetic configurations allowing new sounds and musicality to emerge. There
26
Auditory masking effect occurs when the perception of A is affected by the presence of sound
B. It also concerns simultaneity, frequency, spectrum and directional masking
14
is hopefully more to explore in order to create music with new sounds, finding their
own expressiveness. There is also humor in there, like the grotesque walk of the little
soldiers, both burlesque and tragic, ridiculous and terrifying at the same time.
It is difficult to find the right balance between movement and immobility. Like watch-
ing a dancer on stage, and appreciating the choreography, it is easier if the stage and
the spectator do not move too much themselves. Here there are sometimes too many
unnecessary swirls, but I have a hard time deciding which ones to cut in order to stop
the movement in the right place, moment and positions. Two approaches are possible:
either :- i) starting from fixed sounds and building the movement at the time of the
compositional process, which means adding a disconnected movement on top of an
existing playing sample ii) or pre-constructing the movement inside the sound materials
themselves and then reducing or fixing them afterwards, with the inconvenience that
the movement is fixed and attached to the sound. I mostly used the second method.
I have moved on from the idea that sounds have to be noisy or contain noisy attacks to
be precisely positioned, while tonic and sustained sounds are difficult to locate. That is
still true, but intermediaries also work. It is an important compositional choice that the
positions be clear and precise for some sounds, less for others. Trajectories require a
fairly high spatial precision, unless listeners simply feel that spatialisation has changed,
but without perceiving kinetic qualia. Spectral, temporal, pattern complementarity is
also a major tool for distinguishing kinetic qualia. For example, impulses are very easily
positioned on top of sustained sounds and seem to appear at the forefront. This com-
plementarity is a key to access musical complexity.
Kinetic design is a composition that must be listened loud enough but not to the point
of needing ear protectors. Physiologically speaking, it is very demanding for the lis-
tener’s ears. At the end of listening to it, listeners may start to feel this composition has
made them work in various unusual registers but without pain or hearing fatigue.
6 Conclusion
This design process has proved efficient in a number of ways. A large part of the audi-
ence who listened to Kinetic Design said they had a singular experience of spatialisa-
tion. For the first time, they could really hear the movement of the sound. The spatiali-
sation was effective and impressive. They also said that it was a show of sound, the
plasticity of the sound was expressive and constituted a novel listening experience.
However, does this mean that there is a new musicality, a kinetic music, a new genre
in music? The answer is more nuanced. Some other composers working in the same
field who listened to the performance were quite critical and said that the music did
move but they could not feel any particular significance in the movement, nor a specific
kinetic musicality. Any evaluation from listeners’ experience is only representative of
a sample; it is bound to be subjective and must therefore be relativised and treated with
caution. The question whether kinetic music can be developed into a new expressive
genre or is only a fictive construction to push spatialized expressivity forward to its
limits, still remains of interest and needs to be explored further with more
15
experimentation, applied research and experimental creation. New tools and design pro-
cesses also may help to improve kinetic music experience.
References
1. Baalman, M.: M.A.J. 2008. On Wave Field Synthesis and Electro-Acoustic Music, with a
Particular Focus on the Reproduction of Arbitrarily Shaped Sound Sources. PhD thesis, Tech-
nische Universität Berlin
2. Baalman, M.: Spatial Composition Techniques and Sound Spatialisation Technologies, Or-
ganised Sound15(3): 209–218 & Cambridge University Press, 2010
3. Blauert, J.: 1997.Spatial Hearing. Cambridge, MA: TheMIT Press.
4. Bregman, A.: A.S. Auditory Scene Analysis: The Perceptual Organization of Sound. Cam-
bridge, Mass.: Bradford Books, MIT Press (1990).
5. Cahen, R. : Octophonie DeltaP-Cahen. Compte rendu de recherche pour Octophonie Studio
Delta P La Rochelle (1994).
6. Hauberg, J.: Research by Design – a research strategy
7. Macedo, F.: Phenomenology, spatial music and the composer: prelude to a phenomenology
of space in electroacoustic music: Proceedings of the International Computer Music Confer-
ence 2011, University of Huddersfield, UK, 31 July - 5 August
8. Merlier, B.: Vocabulaire de l’espace en musiques électroacoustiques DELATOUR France,
pp.230, 2006. Hal shs-0051174
9. Pottier, L. La Spatialisation Des Musiques Électroacoustiques. Université Jean Monnet-Saint-
Etienne. Edited by C.I.E.R.E.C Travaux ; Musique et Musicologie 157. Saint-Étienne: Publi-
cations de l’Université de Saint-Étienne, 2012.
10. Schaeffer, P. : De l’expérience musicale à l’expérience humaine. La Revue musicale n°274-
275. Paris, Richard-Masse (1971).
11. Schaeffer, P. : Traité des objets musicaux essai interdisciplines. Paris: Éditions du Seuil,
1977.
12. Schaeffer, P.: Machines à communiquer Tome 1 Genèse des simulacres
13. Thigpen, B.: Spatialization Without Panning. eContact! 11.4 Symposisum électroacoustique
de Toronto CEC https://econtact.ca/11_4/thigpen_spatialization.html , last accessed
2019/04/20 (2009).
14. Verron, C. : Synthèse immersive de sons d’environnement. Thèse de doctorat en acoustique
Université Aix-Marseille (2010).
