Sonic Composition in "Tongues of Fire"

Author(s): Trevor Wishart

Source: Computer Music Journal, Vol. 24, No. 2 (Summer, 2000), pp. 22-30
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/3681924 .
Accessed: 10/07/2011 09:22

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=mitpress. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to Computer Music
Journal.

http://www.jstor.org
TrevorWishart
Department of Music
Sonic Composition
University of York
Heslington, York YO1 5DD, UK in Tongues of Fire

Editor's Note: The sound examples cited in this manipulation of a sound to produce related
article can be found on the Computer Music Jour- sounds, while the term transformation refers to a
nal CD volume 24 that accompanies the Winter process of sonic development through time; that
2000 issue. is, the use of sonic relationships between events to
build musical structures in time. This is also re-
In composing, I differentiate between the proce- ferred to as sonic modulation by analogy with mo-
dures I use to generate compositional materials tion between keys in the tonal system. We can
and the structures I lay out for the listener. It's not also proceed in several dimensions at once, and in
important for the listener to know anything about fact the discrete parameterization of sounds in
my compositional procedures, but I hope they in- terms of pitch, duration, etc., does not sit easily
form the formal and dramatic structure that I do with our perception and recognition of sounds in
hope the listener will perceive and appreciate, the real world.
though not necessarily consciously. Hence, much The method I therefore adopt is quite similar to
of this article is for the benefit of composers rather the computer-graphics model of evolution pro-
than potential listeners. posed by Richard Dawkins in The Blind Watch-
My composition Tongues of Fire (1994) is a mu- maker (1986): I begin with a starting sound, and
sical work created for the recorded medium using apply any number of different small metamorpho-
the computer. It relies on the computer's signal- ses. A metamorphosis must lead to a perceptually
processing power to metamorphose one kind of similar sound, as discussed later. I then transform
sound material into another, thereby making au- many of these sounds a little, and so on. In this
dible connections between different kinds of way, I build a tree of interconnected sounds with
sounds and enabling a musical structure to be de- its nodes and branches. As this process of sound
veloped in the sonic domain. generation proceeds, I select particular sounds to
Because sonic space is multidimensional and the further metamorphose, or to use finally in the
choice of possible starting sounds is unlimited, we piece, on the basis of their intrinsic aesthetic
need some way to navigate through this fascinat- qualities and their audible relationships to one an-
ing space of possibilities. In the early- and mid- other. However, some sounds may be used purely
20th century, composers became transfixed by because they form a perceptual bridge between
permutational procedures, deriving ultimately two other more notable sounds, like passing tones
from the serial method of Schonberg. However, in a pitch-based musical organization.
these procedures relied on a finite set of discrete This tree-generation approach is a generative
elements to permute. In contrast, sound-space is method, not a compositional rationale. In many
unbounded and continuous. A more detailed dis- cases, sets of related sounds used in the piece may
cussion of these ideas can be found in my books be derived from connected points along the tree,
On Sonic Art (Wishart 1985) and Audible Design but this is by no means necessary. Ultimately, my
(Wishart 1995). perception of the relationships among sounds de-
We can proceed seamlessly from one point to an- pends on my listening to them, and when I make
other (a continuous transformation) or by discrete such judgments, I have no interest in how, or even
steps (a sequence of discrete metamorphoses, more whether, the sounds are generatively related. I have
akin to traditional motivic variation, perhaps). I criticized elsewhere the confusion of composi-
use the term metamorphosis to refer to the sonic tional methodology with audible design (see, for
example, chapter 9 in Audible Design). Sounds
Computer Music Journal,24:2, pp. 22-30, Summer 2000 originating from very different generative processes
? 2000 Massachusetts Institute of Technology. may have significant audible similarities, while

22 Computer Music Journal

sounds from the input and output of some process at any stage of sequencing or layering, can itself
may have no notable audible similarities at all. become the subject of further metamorphosis.
In practice, however, the generative processes I The "theme" of Tongues of Fire is a rapid solo
use (and which I have developed) tend to produce vocal utterance less than 2 sec in length. This
sounds that are perceptually related to each other theme was chosen both for expressive reasons (it is
in some way. Otherwise, the generation of mate- recognizably human, but slightly grotesque,
rial would become arbitrary,and the search for slightly comical, and without any linguistic con-
connections hopelessly complex. Hence, the pro- tent in any existing human language) and for sonic
cess of evolving materials is not strictly random, structural reasons (it is a sequence of several spec-
as in Dawkins's sense. I use the analogy to stress trally complex and different sounds, thus making
that it is also not deterministically rational. Part excellent raw material for many kinds of sonic
of the excitement of composing with sounds in metamorphoses). The theme of Tongues of Fire
this way is the often-unpredictable consequences can be heard, repeated, in Sound Example 1.
of applying metamorphic processes to new types of The theme is immediately repeated. Structurally
sounds-consequences to which one must be con- speaking, I tried to make it clear to the listener
stantly open. that this surprising and-on first hearing-appar-
This quasi-genetic process of variation and se- ently arbitraryevent is significant for the piece. In
lection combines rational exploration with per- fact, all of the sound material in Tongues of Fire
sonal aesthetic choice. This balances two develops from this seed. In this sense, Tongues of
important aspects of composing for me: the things Fire does not differ from a traditionally conceived
that we know that we know (explicit perceptual instrumental work that proceeds by the metamor-
connections and rational processes of metamor- phosis and development of notated pitch motives.
phosis and transformation), and things that we However, Tongues of Fire deals with sonic meta-
don't know that we know (cultural and personal morphosis and development. Using the computer,
implicit preferences that may be far more interest- we can apply many processing procedures both di-
ing to future listeners than all our intellectual ra- rectly to the waveform of the sound or to the time-
tiocinations, because they carry information about varying spectrum of the sound, and we can
cultural context and our relationship to it that we reassemble the sound in many different ways, at
have simply taken for granted). many perceptually different timescales (from
This process forms the basis for the next stage, phrase frame to rhythmic frame to grain frame to
in which the materials are arrangedin temporal imperceptibly short sonority frame). Further dis-
sequence, or overlaid contrapuntally (strictly cussion of this procedure may be found in Audible
speaking, as counterstreams) to emphasize, mask, Design (pp. 16-19).
or complicate their perceptual connection, form- In this sonic space, traditional categories and hi-
ing an evolving structure in time. It is primarily erarchies of musical parameters break down. In
this temporal unfolding of the materials that I particular, "timbre"-essentially an umbrella term
hope the listener will perceive and appreciate. for all those features of sounds not captured in tra-
In reality, there is not a simple dividing line be- ditional notation, including stationary spectra,
tween the two stages. Often, sets of metamor- spectral changes, formant structures, formant
phosed material are worked into complete phrases changes, vibrato/tremolo and their evolution,
at an early stage. Some continuous metamorpho- graininess, jitter, etc.-provides a multidimen-
ses call out to be treated as musical phrases in sional arena in which pitchness is but one dimen-
their own right, as they extend over a considerable sion. We can navigate and explore the timbre of
time (for example, the voices to water transforma- sounds in detail for the first time, now that com-
tion discussed below, lasting for around 90 sec) puter analysis and manipulation provide us with
and provide a sense of musical motion within the power to do so. (Pitchness and pitchal are
themselves. Also, any concatenation of materials, terms used in Audible Design to designate the

Wishart 23
spectral quality of a sound that gives it a clear the sound to be). No matter how abstracted from
sense of possessing a single pitch, as opposed to be- reality our sounds might be, we will still tend to
ing noise-like, inharmonic, or multipitched. The attribute material and dynamic causation to them
term is used to distinguish it from the harmonic that is based on our experience of sound in the real
sense of pitch. Thus, a rapidly sliding sine tone, es- world. (See for example the chapter entitled "Is
pecially one that crescendos from inaudibility and There a Natural Morphology of Sounds?" in On
then diminuendos into inaudibility, may never Sonic Art.) Hence, sonic metamorphosis can be a
have any specific pitch in the harmonic sense, yet multifaceted and subtle art.
it is perceived as having a pitch-like spectral qual- However, occasionally I allow myself to "run
ity, to have pitchness, or to be pitchal.) with" the result. For example, just beyond the
The signal-processing power of the computer pulsed-rhythmic climactic section of the piece (at
means that the starting sound becomes almost in- ca. 20:20), there is an extreme metamorphosis. I
finitely malleable. In these circumstances, we usually refer to this as the "fireworks" metamor-
need to make a distinction between what is tech- phosis. I always attach names to sounds, giving
nically a metamorphosis and what is perceptually some hint of their audible substance. This, for me,
a metamorphosis. As an extreme example, any is a vital organizational procedure when dealing
starting sound can be converted to noise by multi- with hosts of sonic material. Simple numbering or
plying each successive sample by a different ran- technical naming makes it difficult to locate a
dom number. Technically speaking, this could be sound again when one has a musical, rather than
regardedas a metamorphosis (the algorithm in- purely technical, goal in mind. One advantage of
volved is quite straightforward),but perceptually, composition with computers is that one is forced
there is no relationship between the starting sound to give a name to all soundfiles, whereas in the
and the resultant sound (and in fact the same or analog studio, one is faced with segments of tape,
very similar perceived resultant could arise, re- looking more or less identical (apartfrom their
gardless of the starting sound). However, even in length), and elaborate cataloging procedures must
this case, it is possible to conceive of creating a se- be followed to keep track of sound materials.
quence of increasingly noisy versions of the origi- The immediate starting sound (or sonic node) for
nal sound, spanning the timbral space between it, this metamorphosis is the sound voismetal (dis-
and noise (compare, for example, the metamorpho- cussed in more detail later), which has a percussive
sis of the boy's singing voice to a pure sine tone in vocal attack leading to an extended "metallic"
Karlheinz Stockhausen's Gesang der Jiinglinge inharmonic sound whose tessitura descends
[1958]). Perhaps in this way we might create a set slowly. The metamorphoses use waveset averaging:
of perceptually acceptable transformational steps the duration and shape of the starting sound's wave
for the listener, linking starting sound to noise. between each zero crossing is averagedover time.
Moreover, when presenting this material in a Applied to such a complex starting sound, waveset
piece, we don't need to sequence the increasingly averaging produces many unpredictable and noisy
noisy sounds in that temporal order. As long as it artifacts. The only perceptibly retained feature of
is possible for the listener to make the connection the starting sound is the descent in tessitura. Al-
at some stage of a repeated listening, we have a though it is on the edge of perceptual justifiability,
perceptible transformational structure, an audible I felt that this extreme metamorphosis fit well at
connection between two quite distinct sounds: our this particularly dramatic moment in the piece,
starting sound and noise. where the energy of the pulsed rhythmic material
In practice, I am more perceptually demanding subsides. Here, the piece has reached its transfor-
than this example suggests. Quite subtle changes mational outer limits, and is about to tie up the
in the nature of a sound (for example, changes to loose ends in a traditional coda, using mainly reca-
the attack characteristics) can radically alter our pitulated (rather than further metamorphosed) ma-
source attribution (what we imagine the source of terials from throughout the piece. The

24 Computer Music Journal

time-stretched voice and the fireworks metamor- an inharmonic, semi-metallic quality. This tail
phoses can be heard in Sound Example 2. glides down slowly in tessitura. This vocal-attack-
The larger-scale structure of Tongues of Fire to-metallic-extension sound is the starting point
does not differ markedly from that of a comparable or sonic node for many sonic processes, and I refer
instrumental work. At the largest time frames, we to it as voismetal.
are no longer dealing with source recognition or The second theme variant is a texture of
sonic detail, but with overarching connections be- gabbling voices, extending the solo voice theme
tween blocks of material. Hence, in some ways, material into a disgruntled-crowd-like event by su-
there is no reason to abandon existing procedures, perimposing several different variants through
though of course new means of formal organiza- time. I refer to this as gablcrowd. This variant is
tion become possible, particularly those using also an important sonic node.
source recognition to project overall sound land- Last of all, the percussive vocal attack men-
scape or narrative. tioned before is stacked (several tape-speed-like oc-
Just as the initial theme is immediately repeated, tave transpositions superimposed) with attacks
the whole first extended phrase, in which that synchronized, and a kind of reverberant extension
theme is developed in a number of ways, is followed added to the higher-and hence shorter-compo-
at ca. 1:00 by an echoing phrase, the theme now de- nents. This brings out the pitch (D) of this event,
veloping differently. Similarly, on an even larger which otherwise is difficult to focus upon in the
scale, the first major section ends at ca. 10:00, when sonic context of the theme. This variant appears
the progressive granulation of the voice slows to a very close to the start of the piece, and its
regular clock-like tick, leading to a pause. After the pitchness sets it apart in the nonpitchal context. I
pause, the theme returns but is then followed by refer to it as pichstak. It also becomes the seed of
four cycles of a rhythmic variant of that theme, further developments.
which becomes crucial to the continuing evolution The original theme, its rhythmic variant,
of the piece. It recurs at 11:50, 13:10, 16:50, and voismetal, gablcrowd, and pichstak can all be
17:00 (where it already takes on some of the pitched heard in Sound Example 3. The development of
characterof the later development). This leads ulti- the rhythmic version at the climax of the piece
mately to the material at 18:20, where it forms the can be heard in Sound Example 4.
rhythmic basis of an extended pulsed-rhythm sec- Let's now return to the opening of the piece.
tion in which changing pitchfields become impor- Sonic development begins immediately. The vocal
tant, and which forms the climactic conclusion of theme itself starts with a percussive vocal attack,
this part of the piece. After the dying away of activ- and ends with a "slurp" sound. These elements are
ity (from 20:40 to ca. 22:20), short segments from all immediately developed. After the two statements
over the piece are juxtaposed amid silences at the of that theme, a third statement begins with the
start of the coda. At the end of the coda, the theme percussive attack, but is truncated directly to the
is recapitulated in a truncated version of the open- slurp, which is then repeated, semi-overlaid, at
ing. The piece concludes with a fragment of the lower pitches and speeds, leading to the pichstak,
theme as a cadential event. the strongly pitched version of the percussive. One
Three other theme variants play an important of the few steady-pitched sound events in the
role in the piece. The first is a metamorphosis of piece, pichstak becomes an important marker, an-
the almost percussive vocal attack of the theme, nouncing the start of short or long phrases, some-
in which the tail of the sound is greatly stretched times transposed, until it becomes the pedal point
in time. (The initial attack is not stretched, in or- of the harmonic sequence of the pulsed-rhythmic
der to preserve the sense of a vocal origin for the climax previously mentioned.
sound.) With extensive stretching, the noisy spec- At its first occurrence, pichstak introduces a de-
trum, a spectrum that varies rapidly in a brief celerating sequence of the original percussives.
space of time, becomes slow moving and takes on This deceleration prefigures the accelerating

Wishart 25
"bouncing" sounds to come. The deceleration is seamlessly becoming something different, like the
achieved with a nonstandard time-stretching tech- voice-to-bees metamorphosis in VOX 5 (Wishart
nique, giving stretched duration proportions of 1990), or as a sequence of discrete steps, like the
1:2:3:4:5:6:7.The stretching process divides the metamorphosis from singing voice to pure sine tone
original signal at every other zero crossing, and previously referredto in Gesang der Jiinglinge.
treats the signal segments thus produced as indi- In the continuous case, and also where discrete
vidual waveforms. Time stretching is achieved by metamorphic steps lead gradually away from one
the immediate N-times repetition of each wave- type of sound to a recognizably different type, I
form before proceeding to the next. Clearly, with a have compared this sonic process to key modula-
simple waveform such as a pure, fixed-frequency tion in the tonal system. Clearly, sonic space does
sine tone, this procedure produces perfect time not have the strictly symmetric and cyclical struc-
stretching, but applied to complex or rapidly ture of the tempered scale's cycle of fifths. We do
changing signals, it generates strange artifacts. In not even have an analogy of the octave in sonic
particular, once a waveform has been repeated space. But it is possible to create a similar sense of
about four times, it will generate a unique pitch. movement from one part of musical space to a rec-
(This is based on my own experience. It may be ognizably different one, for example, from voice-
that more cycles are needed with very high fre- like to wood-like, or to return to a sound sonically
quencies.) It is therefore possible, for example, to close to the original starting sound and to have a
start with a complex signal and proceed, step by sense of "distance" from our starting sound. Thus,
step, to a 256-times repetition of each component just as F-sharpmajor is much further from C major
"waveform," thereby generating a melodic se- than is G major, so an inharmonic resonant bell
quence of arbitrarytimbres and arbitraryrhythms. sound is further from the sound of pouring sand
In our case, as the number of waveform repetitions than from an inharmonic nonresonant bell sound.
increases, the original signal gradually breaks As sonic space is multidimensional (Wessel 1979),
down into an ultra-rapid stream of brief pitch it is difficult to impose any simple measure of sonic
cells. Toward the end of our sequence, therefore, distance (a metric) on that space; whereas in tonal
this pitch bubbling becomes apparent. space, G major is the same distance from C major
Near the end of the sequence, the same percus- as it is from D major, measuring around the cycle
sive vocal sound reappears, grouped into an accel- of fifths. However, I am not concerned with draw-
erating and decrescendoing set, suggesting a ing an exact parallel, but rather with merely dem-
bouncing object as its source. This rhythmic/loud- onstrating a certain analogy in the sonic domain
ness motive is repeated with sonic variations in with tonal motion, tonal return, and tonal distance.
which the waveforms are altered to a different We begin by examining the phrase that starts
shape for each repetition of the motive. Perceptu- with the pichstak D attack at ca. 2:40 and lasts un-
ally, we seem to hear bouncing on or in different til ca. 4:00 (as in Sound Example 6). The node
physical materials, for example, sand. The whole sound from which this phrase develops is
phrase concludes with a repetition of the resonant voismetal, and is heard immediately after the
D, sustained, with a slight tremolo as it decays to pichstak attack. It is developed in various ways. In
nothing. The complete opening phrase, as de- the first case, we produce a rather artificial
scribed above, can be heard in Sound Example 5. tremolo in the tail of the voismetal sound by im-
Hence, sonic development is intrinsic to the piece posing a rapid sequence of brief, shallow, loudness
even within its opening phrase, and we could con- dips. The first of these is heard at 2:40 (where the
tinue to examine the whole piece in this amount of phrase begins), and is the first of a sequence of dis-
sonic detail. However, I intend to look at just a few crete metamorphoses that gradually leads else-
example phrases to illustrate some of the multifari- where: a sonic modulation. Thus, on each
ous sonic processes used. Note that a sonic meta- recurrence of the voismetal sound in the ensuing
morphosis may proceed in the continuum, a sound overlayed texture, these loudness dips cut deeper

26 Computer Music Journal

into the sound, and the tail begins to separate per- by waveset distortion. In this process, the signal is
ceptually into a sequence of discrete struck-wood divided at every other zero crossing into wave-
events. (Descriptive adjectives like "metallic," forms, and these waveforms are replaced by differ-
"struck wood," or "cicada noise" are meant to ent waveforms but of the same duration and
suggest the type of sound event being referred to, amplitude. We can then apply a process of in-
rather than to give an accurate description of their betweening, a term I have borrowed from cin-
physicality.) Using a grain-detection program, we ematographic animation, where principal picture
can isolate these brief events and duplicate and/or frames are drawn by the chief artistic designer,
transpose them, and this approach is used to pro- then the many in-between frames that produce the
duce a further variant that more strongly empha- sense of movement linking one key frame with an-
sizes the perceptual distinctness of the other are produced, originally by assistants, but
struck-wood events. By now, we have sonically more recently using computerized methods. In the
"modulated" from stretched voice to struck-wood- sonic case, beginning with the pure struck-wood
like (ca. 2:50). The metamorphosis from voismetal event, we produce a sequence of sounds in which
to wood can be heard in Sound Example 7. more and more of the new waveform (and hence
From here starts a texture in which time-variable less and less of the original "wood") is mixed. As
pitch shifting is used to create variants of the origi- the two waveforms being mixed are synchronous
nal voismetal sound, which pitch-slide down or up at their zero crossings, we are effectively generat-
over largerranges, or which oscillate wildly from the ing a gradual sequence of metamorphoses of the
mean pitch. In the context of the piece, I hear these waveform itself, and hence of the sonority. Putting
more as variations on the voismetal original than as these mixes together in a progressive sequence, we
true metamorphoses of that original, but there is no produce a sense of sonic motion from struck wood
sharp dividing line between the two, and to some ex- to "drum-like." The sequence also accelerates,
tent it depends on where the time sequence of vari- like the bouncing sound referredto earlier, but
ants, as laid out in the piece, leads. In a similar way, now the sound sequence is a crescendo (rather
a pitch may be judgedto be a chromatic passing note than a diminuendo), suggesting, perhaps, inten-
or a pivot of tonal modulation, dependingon what tional force, rather than the energy dissipation of
eventually happens to the surroundingsequence of free bouncing (as in Sound Example 10).
pitch events. This texture forms a bridge to a final The phrase terminates (from ca. 3:52) with yet
strong statement of the struck-wood sequence, the another variant of voismetal. This applies the
separateness of the wood events being even more brassage technique, as used in the "harmonizer,"
emphasized by the ritardandoof the wood sounds (as to voismetal. Standardharmonizer brassage chops
in Sound Example 8). up the starting sound into segments about 50 msec
Here the wood sounds accelerate and rise in pitch, in duration, sufficient to preserve the instanta-
leading into a texture of higher-pitched, shortened, neous pitch of the event, but insufficient to give
repeated sounds, over a small pitch range, vaguely much of a clue to any changes in pitch that might
like cicada noise, an even more distant sonic modu- be taking place. By splicing these segments back
lation (3:07-3:17, in Sound Example 9). together again, in the same order but with a
We also hear here a dramatic extension of the greater or lesser degree of segment overlap, we can
voismetal sound, a stack of transposed versions of reproduce the original event with an altered dura-
it. This is also prominent elsewhere, and it serves tion, but with no shift in pitch. (In reality, we
as the penultimate sound of the entire piece. This have to slightly randomize the durations of the cut
then becomes the second part of a repeating phrase segments to avoid creating a spurious pitch from
unit (3:17 and 3:28), whose first part is an acceler- any regular sequence of splices used.)
ating sequence of attacks. This accelerating mate- This process can be varied in many ways. In par-
rial also emerges from the struck-wood events. ticular, we can adopt the following routine. First,
The waveform of one of those events is modified we cut up the original sound as usual. Then, for the

Wishart 27
first segment of the reconstructed sound, we choose superimposed shreds, the shredded gablcrowd
the first segment of the original sound, but for the sound retains some vocal clues. Eventually, after
second segment of the reconstructed sound, we thousands of shreds, the dominant events in the
choose the second or the first segment of the origi- resulting sound will be the splices themselves, and
nal sound. For the third segment of the recon- any originally complex starting sound will be re-
structed sound, we choose the third or the second, duced to low-level noise. However, there is a point
or the first segment of the original sound, and so on. along this path of successive shreddings where the
As a result of this selection procedure, there is vocal quality of the starting sound is lost, but not
always a possibility that some of the early, percus- its sonic diversity. At this stage, the sound be-
sive-attack segments of the original sound will ap- comes akin to the sound of water falling gently
pear anywhere in the reconstructed sound. Hence, around stones in a mountain stream.
this attack quality gets distributed randomly over Our phrase (13:20-15:10) is constructed by splic-
the time-stretched resultant sound, producing the ing together successive, cumulative shreds of
"gargling"quality that we hear after 3:52 (as in gablcrowd so that the voices gradually dissolve
Sound Example 11). into the water-like texture. This basic phrase
Thus, by putting all these variants of the structure is counterstreamed by other events: deep
voismetal sound node together in an appropriate attacks initiating rising noise-bands derived from
temporal sequence, we have constructed a com- the shredded voices. It also continues, seamlessly,
plete phrase based on the sonic development of the into a second continuous metamorphic process,
original sound node (as in Sound Example 12). evolving a pitchal upward portamento out of the
This is an example of phrase construction mainly noise texture to lead us into the pitch attacks of
using discrete sonic metamorphoses. However, we the next phrase.
can also generate entire phrases by the continuous The technique used here is end-synchronized de-
or semi-continuous metamorphosis of the starting lay. First, several copies of a sound are made at
sound. In the two examples considered here, we slightly different speeds, analogous to tape-speed
are using gablcrowd as the sound node from which variation. For example, we might decimate a wave-
to begin. form of 48,000 samples to 47,600 samples, thereby
The first example applies sound shredding to changing both duration and pitch very slightly.
gablcrowd. The shredding process takes a given These copies are then superimposed so that they
duration of sound material, and cuts it at a given synchronize at their beginnings. The resulting pre-
number of randomly chosen positions; for ex- cise beat frequencies caused by the precise delay
ample, seven cuts divide the sound into eight dis- times between the copies will be heard as a pitch
junct segments, each of random length, though the portamento descending from "infinitely" high at
sum of these lengths equals the original duration. the outset, where the copies are all synchronous,
These segments are then shuffled into an arbitrary through the audible-pitch range as the copies get
order, and rejoined to produce a new sound of ex- out of phase and the increasing delay gap generates
actly the same length as the original. This process a falling pitch, into the sub-audio range, resolving
is then repeated many times. In general, at each into a sequence of decelerating echoes. If instead
shred, the new cuts are unlikely to coincide with we synchronize the ends of the copies, we will be-
any existing cuts, so any preserved continuous gin by moving into accelerating echoes, leading
stretch of the original sound will tend to be cut into a rising-pitch portamento. This is what we
again into even smaller segments. hear at our phrase end. This effect occurs no mat-
At each shred and shuffle, the starting sound is ter what the nature of the original sound: it is a
increasingly fragmented and increasingly ran- process-determined (rather than a source-deter-
domly shuffled. Of all our sound experiences, vo- mined) effect, so we can use it to move from
cal sounds retain their recognizability, even under nonpitchal sounds into pitchal sounds. This entire
extreme deformations. Thus, even after a hundred phrase can be heard in Sound Example 13.

28 Computer Music Journal

 As a second example of phrase structure based on ity. As the evolving loudness and timbral envelope
a continuous metamorphosis, we again start from of a sound is often a vital recognition cue, revers-
gablcrowd at 17:07. The metamorphosis begins by ing it can completely alter the perceived quality-
making the crowd texture increasingly dense, by and hence the recognition characteristics-of that
superimposing randomly delayed, inexact copies of sound. In general, however, continuous sonic
the starting sound. The copies are not exact copies, metamorphoses can be retrogradedin this way
to avoid artificial-sounding delay or pitch effects without too many surprises. But if our phrase
between copies. With complex material like this at structure is built out of discrete metamorphoses,
sufficient density, the spectrum becomes satu- we must use the event-order reversal form of retro-
rated, and we hear just a band of noise. At the same grade to achieve a comprehensible sense of retro-
time, this band is made to rise in tessitura. As the gression. Or we may need to use some
process evolves further, the original gabbling combination of the two.
voices re-emerge occasionally, retaining the con- In the next example (7:40-8:30), we begin with
nection with the starting sound. The noise is next gablcrowd, raising the tessitura and then applying
put through a filter bank of increasing Q prior to spectral tracing to the sound. Spectral tracing pre-
the application of portamento, so that it gradually serves the most prominent partials in the sound,
becomes a set of parallel portamentoing pitches. moment to moment (window by window). With
The continuing portamentoing means that the simple, unchanging sounds, this process produces
"chord"is never able to settle in any particular first an elementary noise reduction and then a
harmonic space (therefore it is not a chord in any gradual simplification or impoverishment of the
traditional sense, technically speaking). The phrase spectrum. With complex sounds, however, the set
continues by reversing these processes, returning of N most prominent partials changes on a mo-
first to noiseband and then back to gablcrowd. A ment-to-moment basis; some partials leave the set
counterstream of portamentoing parallel pitches while some enter, resulting in the revelation of
splits away during the course of the phrase and de- complex weaving "melodies." (An example of
scends slowly in pitch, in contrast to the up-and- spectral tracing can be heard in Sound Example
down wave-like motion of the principle stream. 15.) In this example, the originally vocal texture
This phrase is thus based on a classical arch form gradually becomes pitchal in the timbral sense,
using continuous sonic metamorphoses (as in rather than noise focused. After somewhat mecha-
Sound Example 14). nistic echoes, we reach a metallic-clang event gen-
An arch form might be regardedas a sequence of erated by stacking transposed copies of
events or sounds followed by its retrograde.The exponential decay-enveloped versions of the spec-
idea of a retrograde,however, is not as straightfor- trally traced sound (as in Sound Example 16).
ward as it seems. In traditional notated music, a When this material is later recapitulated in ret-
retrograde of a sequence of musical events means rograde form (10:50-11:40), we hear the continu-
that we play the original events in reverse order. ous metamorphosis evolve in truly reversed time,
We are retrogradingonly the starting times of the from weaving melodies to transposed gablcrowd to
events. We do not reverse the time flow inside the original gablcrowd. However, the metallic-clang
events themselves. event is presented again in its original, time-for-
We could extend this notion of retrogradeto dif- ward form, and it subsequently develops through
ferent time scales; for example, a dance piece of transposed repetitions of this original form. Fur-
overall form ABBA might be thought of as having thermore, as we already know this material (we
a retrograde structure. However, if we reverse the have heard it previously in the piece), I have
flow of time itself, by running a tape backwards, or placed a counterstream against it. Here, "breath-
reversing the order of a sequence of samples, the ing" sounds shorten and accelerate, becoming
result is often surprising to us, as we reverse the click-like and evolving into a regular pulse, like
events themselves and hence their flow of causal- the "clock" first heard prior to the pause at 10:00,

Wishart 29
and leading into a recapitulation of the pulsed- stands or falls by its impact in the real-time aural
rhythmic version of the theme at 11:45. Further- experience of its listeners.
more, the clock and metallic-clang "bells" are
briefly associated in a tangential reference to some References
external reality-but this brings us into a different
realm of analysis. This process may be heard in
Dawkins,R. 1986. "TheBlindWatchmaker."New
Sound Example 17. York:Norton.
We could continue examining Tongues of Fire Stockhausen, K. 1958. Gesang der Jiinglinge. Cologne:
phrase by phrase, but I hope I have given sufficient Stockhausen-Verlag.
insight into the musical processes at work. I have Wessel, D. 1979."TimbreSpaceas a MusicalControl
concentrated on the use of sonic metamorphosis Structure." Computer Music Journal3(2):45-52.
in phrase building, and have said little about crite- Wishart, T. 1977. Red Bird. Albany, NY: Electronic Mu-
ria for counterstreaming or for the more general sic Foundation.
sequencing of events. Nor have I discussed issues Wishart, T. 1985. On Sonic Art, new ed. London:
of sonic landscape, which in Tongues of Fire are Harwood Academic.
dealt with in a more tangential manner than in my Wishart, T. 1990. VOX 5. In The Vox Cycle. York, UK:
Orpheus the Pantomime.
other compositions Red Bird (1977) or Fabulous
Wishart, T. 1994. Tongues of Fire. York, UK: Orpheus
Paris (1999). These are of equivalent importance in the Pantomime.
putting together the piece, but they are probably Wishart, T. 1995. Audible Design. York, UK: Orpheus
easier to relate to traditional musical form-build- the Pantomime.
ing techniques; therefore I'll leave these matters to Wishart, T. 1999. Fabulous Paris. In Or Some Computer
other analytically inclined listeners to tease out. Music. London: Touch.
But in conclusion, I must add that Tongues of Fire

30 Computer Music Journal