LAB EIGHT

REVIEW
Understand volume automation and be able to create volume
automation data.
Understand the following concepts, which we explored
through the various tasks in Lab Six:

variety through reordering

creating longer gestures

varying repetitive material

bouncing gestures to create longer gestures

reversal, pitch shifting, speed change, timbral processing,

delay, and reverb.

MOVING FROM SOUND OBJECT

TO GESTURE
The practical steps involved in creating an electroacoustic
composition are as varied as the styles of music contained within
the genre. There is no set working method; composers will find a
way of working that is successful for them.
In general, you will begin by finding your sound object(s)
(Lab Four), recording and editing them (Lab Five), and experimenting with them through initial processing (Labs Five and Six).
Sometimes, the processes may create longer ideas, or gestures: for
example, by slowing down the material or transposing it down
several octaves. Other times, the processing may suggest other
processes that will work in combination. For example, transposing a
sound object by varying degrees can create both variety (there are
several versions of the process) and unity (they are all transpositions
of the same sound object). Another possibility is to combine similar
processing of different, or related, sound objects. For example, you
might combine various speed changes of several different door
slams. There would be less unity and more variation than if the
doors were all the same. Such continua are important in establishing
an overall sense of variation throughout your composition.

163

Lab Eight

Beginning to Combine Sounds Early

Once you have a number of sound object transformations, you
might want to try combining some of them within ProTools to get a
feeling for how they go together. It is important to experiment with
the sounds in a compositional setting because you can determine
whether your transformations are too unrelated or if you need
more versions of a particular process. For example, after you
combine several pitch changes for a longer sustained sound, you
might feel that you need to create more versions (at different
transposition levels). You might also feel that the one transposition
up four octaves doesnt seem to fit in with the others, and you may
not want to use it after all.
Creating Many Sessions/Projects
Dont feel that you should be using only a single session/project for
this project or for any other composition. Unlike painting, in which
you would most likely begin and end with the same canvas, you
should be creating lots of sessions/projects within which to build
smaller ideas, then you can combine them in other sessions.
Switching between Composer and Sound Designer
Sound design is the process of experimenting and creating the
individual sound objects; composition is the process of putting
these sound objects together. They should not be mutually
exclusive processes; rather, you should be interchanging these
steps. While you are designing your sounds, you might get an idea
about how the sound can be combined with other sounds: open up
a new session/project and try it out. Similarly, while you are
composing, you might feel that a few more variations of a certain
sound object are required. Fire up your audio editor and begin
experimenting. In this latter case, you will find that your
experimentation will be more focused since you will have a clearer
idea of what kind of sound you specifically need.
Working and Thinking in Layers
Consider creating a session/project for every layer so that you can
concentrate upon it in terms of interest, unity, and variety. For
example, if you need to create a rhythmical layer, dont simply take
a sound and duplicate it dozens of times like an incessant drum
beat; the lack of change and variety will quickly tire the listener.
Instead, use a separate session/project to create the layer, and
concentrate on maintaining interest through variation and
surprises, then you can bounce it and bring it into a master
session/project. If each layer can maintain the listeners interest on

164

Lab Eight

its own, it can last much longer, and it will also contribute to the
overall success of the composition.
Referring (Often) to Unit Seven: Compositional Strategies
Finally, reread the unit on compositional strategies (several times),
and try out some of the ideas.

TASK: ABSTRACTING TEXT

The voice is a very good sound producer, capable of creating a
variety of sounds. These include both harmonic and noise-based
sounds (the human voice cannot create inharmonic sounds) and
short percussive sounds, as well as long sustained sounds. The
voice also has a fairly wide frequency range and the ability to
control timbral change dynamically through resonance. For
example, we can sing a note and change the vowel from ah to
oo to ee and so on. Any of these sounds can be treated as
abstract sound objects and be processed in the same way as jingling
keys, a printing press, or a campfire.
Of course, the voice is also capable of expressing text, words
that mean something. Although we can treat words, phrases, or
sentences as abstract sound objects, we should be aware that the
listener will interpret these sound objects in a different way. Our
brains are wired in such a way that they will automatically attempt
to extract semantic meaning from words.
You may have played the childrens game of repeating a word
over and over until it loses its meaning and simply becomes sound
(try saying frog over and over for thirty seconds). Interestingly, it
takes quite a while before we lose semantic interest in a vocal sound.
We can play with this need to understand text in our electroacoustic music by exploring the continuum between understanding
(semantics) and abstraction (sound).
Setting Up

In a new session/project, import track 59 from the Soundscapes

and Sound Materials CD - Dr. King - as a mono file.

Dr. King is a recording from a vinyl disc of an anonymous

preacher giving a rather boisterous sermon.

Separate the first three phrases into regions. Discard the

remaining text. The first region should contain Now I believe
the words of Dr. King (yeah); this will be referred to below
165

Lab Eight

as Believe. The second region should contain If I helped

somebody (yeah); this will be referred to below as Help.
The third region should contain As I travel on (yeah); this
will be referred to below as Travel.
Reordering (Again)

Move the three regions into alternating tracks (one, then two,
then one, etc.) in the following ordering:

Allow for varying amounts of overlap.

Listen to the result.

Through reordering, you have created a gesture that plays

upon repetition. In a typical sermon, there is always a certain
amount of repetition, so playing on this element is an interesting
possibility. The resulting gesture is different than it would have been
if the source material had been something other than text, since this
gesture could continue longer because of its semantic content.
Varying Spacing and Volume Envelopes (Again)
In order to extend this session in length, we would have to vary the
times between regions more, perhaps introducing silences as well.
One idea that could be explored is the direct repetition and overlap
of regions. Introducing volume changes between regions could
heighten this interplay, and we could perhaps extend the overlap to
five or more regions.
Another possibility would be to play with the background
yeah or the audience response, separating it as a distinct region,
and building upon it by overlapping it with itself (as we did with
the preachers voice).
Processing Text
The next possibility would be to process the text.

Select the sixth region (the third time Believe is used), and
reverse it.

How does this change affect your perception of the continuity

of the text?

166

Lab Eight

Select the tenth region (the fourth time Travel is used) and
pitch shift it (using time correction so that it does not increase
in duration) down three semitones.

How does this change affect your perception of the continuity

of the text?

TASK: COMBINING SIMILAR TRANSFORMATIONSDIRECTION

Reversal is a simple but interesting process because it has become a
recognizable aural icon. Despite the fact that reversal is an unnatural
technique (unlike reverberation, echo, or even filtering, all of which
can occur in nature), most people recognize a reversed sound for
what it is because of its use in music, film, and other media.
Furthermore, certain percussive sounds, or sounds with fast attacks
and longer decays, have a dramatic quality to them when reversed.
Setting Up

In a new session/project, import track 8 from the Soundscapes

and Sound Materials CD - Glass Smashes - as a mono file.

Process the Files

Create reversed versions of each region, while retaining the

originals.

Choose one of the original/reversed pairs, and place the

reversed immediately before the original.

A reversed region, followed by its forward version.

Solo the track and listen to it.

You can hear the clear symmetrical envelope. However,

because your regions probably dont begin directly at the onset of
the attack, there will be a short silence between the two regions.

167

Lab Eight

Overlapping Regions: One vs. Two Tracks

It is possible to move the regions closer together in order to remove
as much of the intermediary silence as possible. It is even possible
to place the forward version in the above example on top of a
portion of the reversed region, so that they seem to overlap.

Placing one region on top of another.

However, the track still remains monophonic; only a single
audio file can be played in a track at a time. This limitation can be
overcome by using two separate tracks and overlapping the regions.

Drag the forward version to the track below with a slight

overlap, solo that track, and listen to both tracks.

A backward region followed by its forward counterpart.

An interesting new sound object. It isnt quite a gesture yet;
its too short.
Combining Processes

168

Select the forward region and pitch shift it, without time
correction, down seven semitones. The duration of the
forward region should now be longer than before.

Lab Eight

Listen to the two together.

This is now more of a gesture since its duration is close to

2.5 seconds. Lets make it a little more dramatic by adding
reverb to the forward pitch shifted region.

Add reverb (about 2 seconds) to the pitch shifted, forward

version.

Listen to the two tracks.

In the above example, I had to lower the volume for the

second track so that it seemed more equal to the backward sound.

Try similar combinations of pitch shifting, reverb, and

forward/backward combinations in order to get a variety of
related gestures.

TASK: COMBINING SIMILAR TRANSFORMATIONSTRANSPOSITION

Concentrating on a single type of transformation process and
presenting the different versions simultaneously is one method of
creating long, interesting gestures. You already experimented with
this process briefly in the previous lab when you combined an
original audio file with the pitch shifted version that used time
correction.
Setting Up

In a new session/project, import the file ahhh bounce that

was created in the previous lab in the Bounce a Gesture to
Create a Longer Gesture task. I hope you still have it; if not,
recreate it following the instructions in Lab Six.
169

Lab Eight

Process the File

Pitch Shift the imported file, using time correction, in the

following ways:

+2 semitones

+3 semitones

-1 semitones

-3 semitones

-6 semitones

Notice that the processes are limited to six semitones (half an

octave) in the lower direction, while the upper end is limited to
three semitones. This will reduce the monster and chipmunk
effects in the vocal transformations.
Also notice that selecting time correction will ensure the
resulting audio files are all the same length. So you will be able to
place them into successive tracks and maintain synchronization.

Place each region into its own track (you will need at least six)
so that they all start at time 0, then duplicate each region so
that the entire gesture is twice as long.

Create a unique and slowly evolving volume automation

envelope for each track, varying each one in a quasi-random
fashion.

Listen to the session.

The result should be a shimmering, evolving chord suggesting

a surrealistic choir. The result shown above is less than twelve
seconds long; a complex gesture like this could, with proper
attention to detail and careful variation, last much longerperhaps
several minutes!

TASK: CREATING AND VARYING A RHYTHMIC LOOP

I am often asked if students should avoid creating rhythmic
material in this course. My answer is always: No, rhythm is
good. I assume that the question results from the fact that most
music played in the class lacks any repetitive, beat-oriented
material. This is mainly because a great deal of electroacoustic
music comes out of the European concert music tradition in which
repetitive rhythm was consciously avoided at all costs. Today,
however, many composers of EA listen to, and are as influenced by,
jungle, drum and bass, and ambient music as they are by the music
170

Lab Eight

of Stockhausen, Subotnick, and Lansky. A rather famous piece by

Aphex Twin, Bucephalus Bouncing Ball, which thoroughly
explores repetitive beats, is undeniably a great EA composition.
The big difference in the use of rhythm between these two
styles is variation. As has been noted at different times in this
course, repetitive rhythm in popular music serves the specific
purpose of propelling the music forward through rhythmic drive. It
is most often a separate layer, and often it is intricately created;
however, once it is established, it must recede into the background
in order to allow other layers, most often voice or melody, to come
into the foreground.
It is perhaps no accident that pieces like Aphex Twins
Bucephalus Bouncing Ball lack text and/or a clear melodic
foreground. Thus, the extra concentration on timbre and rhythm can
be considered more of an experimental electroacoustic aesthetic.
In this task, you will first create a rhythmic loop and then vary
it in a number of ways in an effort to maintain listeners interest.
Setting Up

In a new session/project, import the following tracks from the

Soundscapes and Sound Materials CD as mono files:

Track 39: Chisel 1

Track 47: Wrapping Machine

Listen to the first three seconds of Chisel.

You should immediately notice that there is a distinct

rhythmic quality to this sound. Of course, the rhythm is not
metrically accurate because the sound was not intended to create a
repeating rhythm. However, you are going to extract only a portion
of the sound to create a loop.
Notice that it is fairly easy to visually identify the rhythm
since each stroke of the chisel creates a distinct event.
Highlight four of these events and play the highlighted portion of
the region using Loop Playback.

171

Lab Eight

Select four events for a potential loop.

Notice that the highlight above begins before the first event,
and includes the entire fourth event.
Loop Playback will continuously play and replay the current
selection. Notice that the time counter will continuously cycle
through the selection. When it reaches the end of the selection,
Loop Playback will loop back to the start of the selection,
emulating the tape loop of the classic studio. However, we cannot
use this method to create a loop and add other material because the
session/project never progresses beyond the current selection!
In the case at hand, using Loop Playback will indicate what
the current selection will sound like as a loop.

Play the selection (as a loop).

This could be considered a four-beat loop. Of course, there is

no reason why the loop could not be five, six, or seven beats long;
however, the longer the loop (given this particular recording), the
greater the chance of internal rhythmic inaccuracy.
The loop should sound very smooth as it is repeating. If it
doesnt, zoom in further and adjust the start and end points while
the loop is playing. (You will need to drag the mouse while holding
the Shift key down to alter your current selection).
Notice, however, that the loop will not begin on a strong
event; the sound actually has a noticeable attack. What this means
is that the loop cannot be aligned to other rhythmic material, or any
region with a strong attack (such as a drum-like sound) because the
delayed attack will be noticeable.
For this reason, a loop should most often begin at the greatest
amplitude of the first event. In this case, the last event must then be
subsequently adjusted to include the attack of the fifth event as well.

172

Move the selection over so that it begins at the highest

amplitude of the first event, and adjust the end of the selection
likewise. In other words, the length of the selection must

Lab Eight

remain the same, so check the length of your selection before

you start to move its start and end points.

Beginning the loop on a high-amplitude attack point.

From this selection, create a region. Copy this new region

several times, and place them end to end within a track.
Discard all the extra material.

Turn off Loop Playback and then play the session.

It should sound exactly like the selection did with Loop

Playback enabled. Now, however, time is progressing through the
session, which allows us to add other layers and material as well as
to vary the looped material itself.
Changing Volume Automation
The first process you can explore is simple volume automation. It
can be either detailed to bring out specific beats or more gestural,
using slower rises and falls of amplitude.
Changing One Element in the Loop
One way to add variation is to select a single element and vary it in
some way.

Select the fourth region in the track, and pitch shift it down
one semitone, using time correction to ensure that the
selection remains the same duration.

Listen to the result.

Now you have a slight variation that spices up the rhythm

without disrupting the rhythmic flow. More drastic changes would
obviously add more noticeable variation.

Try randomly selecting regions and transposing them up and

down, both slightly and dramatically. Listen to the overall
result.

173

Lab Eight

Adding an Accent
Accents can be provided using quick volume changes or by adding
in other material in different tracks.

Start again by creating a track with your looped region copied

end to end.

In another track, make a copy of the looped region, and place

this at the exact same time as a region in the next track.

Adding a region to the same time location in another track.

Select this region and pitch shift it down twelve semitones,

(one octave). Again, make sure to use time correction so that
the regions remain the same duration.

Processing an adjacent region.

174

Lab Eight

Play the session.

Notice that because the original loop is still active, the

resulting transposition gives the feeling of an accent rather than
direct variation.
Beat Matching
Beat matching is the process of taking two rhythmic loops and time
stretching or compressing one of them so that their rhythms line
up in time and their durations become identical. This is a very
popular technique that comes out of the techno music of composers
like Ronnie Size, who originally used the technique with digital
samplers to combine multiple drum loops within a single song.

Listen to Wrapping Machine.

Although there is a clear, repeating rhythm, it is difficult to

identify visually because of the constant ambient background noise.
However, if you listen closely, you will notice a repeating double
click, which can be heard as the initial beat in the repeating rhythm.
This double transient is easier to identify visually.

Select a portion of the region between two sets of these double

transients. Make sure to use Loop Playback, and audition the
loop. Adjust the start or end points to fine-tune the loop
length or end points.

A loop in a different recording.

Create a new region based upon this selection, and remove the
extraneous material from the track.

If you compare this loop with the previous, you should

quickly see that these two loops will not work together, because
their lengths are different. This difference in duration is obvious
when you visually compare the two tracks:

175

Lab Eight

Comparing the two regions visually.

This is where time stretching comes in. Use a process in your
multi-track program to alter the length of the region without
changing its pitch.

This is the opposite of pitch shifting with time correction, and is often
called Change Tempo or Change Speed in the Effects menu.

When changing length, you will have to make a note of the

target (destination) length, which in this case is the first loop.
The two tracks should now have regions of identical lengths.

The upper region is time-expanded to match the lower.

176

Lab Eight

Copy this region several times within the track.

Because the regions are of identical length, both tracks should

remain in sync.

Create an interesting volume automation relationship between

the two tracks. Here is one simple example:

A dynamic, rhythm-based relationship between two beatmatched tracks.

PREVIOUS STUDENT WORKS

These are some examples of the musique concrte project from
previous FPA 147 semesters. Presented here in no particular order,
they were the more successful works and so they got the highest
marks. They are given as examples of what can be done by students
within the context of the assignment (rather than in comparison to
the professional works that you have listened to for your Sound
Journals).
The projects are available at the indicated URLs, stored as
monophonic (remember the assignment itself is monophonic) MP3
files. Point your browser to the URLs, and either download the
projects to your computer or listen to them directly in your browser.

EXAMPLE ONE
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_1.mp3
This example opens with two interesting, identifiable sound
objects: a dropping coin and a page turning.
Within the first two seconds, we hear unity (the dropping coin
sound is repeated) and variation (there is a second coin sound).
After several seconds, the page turning sound is introduced.
Whereas the dropping coin suggests boredom (perhaps someone
playing with coins while waiting in line at a cash register), the page
turning is much more frenetic and hurried. Although musique
concrte was not referential (at least in Pierre Schaeffers mind),

177

Lab Eight

these two sounds do have certain referential qualities that cannot

easily be overlooked.
The two sound types (there are variations of the sounds
themselves) continue to be presented simultaneously in separate
layers. Because of their easily identifiable sounds, we, as listeners,
have no difficulty separating the layers in our minds.
Both sound objects begin to be varied through subtle
processing, so the variation still maintains unity: the basic sound
objects remain recognizable. Of particular note is the use of silence
to separate phrases.
My only complaint with the work is the introduction of the
text Figaro. The concept itself is a good one: delaying the
introduction of a sound object until later in the work and then
concentrating on it for the final stage. However, the stylized
singing in a false opera voice tends to cheapen the work and make
it unnecessarily comic.

EXAMPLE TWO
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_examples/
P2_2.mp3
This work has a very interesting opening, beginning with the
evocative high-frequency gesture. In this case, the composer has
started the work with an unrecognizable sound, creating a new
sound world for the listener to explore. There is quite a bit of detail
to the sound, in terms of an evolving, sustained quality that is
combined with a fluttering rhythm.
After ten seconds, the mood is suddenly broken by a
completely contrasting, low-frequency, rhythmic idea. This section
is followed by a voice snippet, and then the low-frequency gesture
is repeated. Contrast is well used here; it is balanced nicely with
repetition (the gesture is played twice), and more variety (the voice
sound object).
The original sustained sound continues with evident
evolutionwe can hear that the sound is slowly, but clearly,
changing and developing. This evolution is balanced with the
outright repetition of the low-frequency gesture, which interrupts
at unpredictable times. Similarly, the voice fragment appears at a
seemingly random rate.
The entire section is interrupted with a new low-frequency
guitar-like sound, which triggers a new text in French. Such use of
178

Lab Eight

a foreign language is a favourite technique of mine, since it is

clearly spoken voice, yet its semantic quality isnt apparent (at least
not to me); therefore, as a listener I can concentrate on the
complexity and beauty of the sound itself.
This text is contrasted with an English passage, beautiful like
they were, an evocative phrase in itself. The following French text
is then cleverly answered by a reversed English passage. We can
recognize that the speaker is using English by the timbre; however,
its reversed nature obviously destroys its meaning and equates it
with the French. Throughout this section, the pitch-shifted guitar
continues in a different layer to add continuity.
The sudden pitch shift in the voice ends this section with
another discontinuity; however, the pitch shift was already heard
in the guitar. The piece ends decisively with the emphatic short
French text.

EXAMPLE THREE
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_3.mp3
This work opens with a nice, long, sustained gesture that comes out
of complete silence, grows to a loud climax, and then quickly
recedes again. This thirty-second gesture is created out of a single
sound: an impressive opening statement. Unfortunately, the digital
artifacts at the climax of the soundthe result of pitch shifting or of
time stretchingare evident, and the somewhat magical quality of
the gesture is destroyed.
The rest of the work continues with long, ethereal, bell-like
sounds that are offset with judicious use of silence and reverberation. Many of the sounds are delicate and very suggestive; however,
their dramatic attacks sometimes cause unfortunate distortion.
Overall, the work creates a fascinating and beautiful sound
world, where objects move and sound with both delicate precision
and fierce authority. I really liked the feeling of the overall work,
which almost convinced me to overlook its technical problems.

EXAMPLE FOUR
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_4.mp3

179

Lab Eight

This is another coin piece, this time exploring many more

concentrated variations of the sound source. Lots of pitch shifting is
used to create different layers; however, the fact that each layer was
derived from the single coin drop gives it a nice sense of unity.
The success of this particular piece lies in its many variations
in processing, with short sounds being contrasted with longer
sounds that are the result of pitch shifting without time correction.
A sense of unity is maintained because a single sound object is used
throughout.
The work has an effective conclusion with the rather long
crescendo: beginning at one minute, the pitch shifted coin roll gets
louder and louder until it is cut off suddenly at 1:28 by the
unprocessed original sound object.

EXAMPLE FIVE
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_5.mp3
This work begins with another unrecognizable sound object,
the chief interest of which is its varying frequency warble. Two
different versions of the sound object are presented for the first
forty seconds. There seems to be a delay process occurring that is
on the verge of feedback; the fast repeated rhythmic element is a
result of this process.
By forty seconds, a new idea appears, although its exact
beginning is hidden. A very low-frequency rumble becomes
evident, together with a contrasting high-pitched wail. The extreme
contrasting frequency bands allow the original sound to continue in
the mid-frequency range.
Again, some unfortunate digital artifacts can be heard,
possibly because of the apparent live delay processing that always
seems to be on the verge of exploding. Again, the resulting interest
in the material almost convinces me to overlook these problems.

EXAMPLE SIX
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_6.mp3
Extreme frequency ranges occur in this work, with lots of
evident pitch shifting. As well, there is an immediate rhythmic loop
which appears and gives the work momentum.
180

Lab Eight

Many of the layers within this work are present throughout its
duration. The layers can be heard through different frequencies,
timbre, and rhythm. For example, notice the extreme low-frequency
rumble; the metallic banging sounds, also pitch shifted; the high,
water-like sounds; and the rhythmic loop. Each of these layers
progresses and develops at its own rate and has internal variation
and interest.
By fading the different layers in and out, the composer gives
the listener something to concentrate on, rather than simply
presenting all layers at the same time throughout.

EXAMPLE SEVEN
http://www.sfu.ca/sca/courses/fpa147/DISTANCE/audio_exam
ples/P2_7.mp3
This is a text-based piece that demonstrates many of the
potential benefits and pitfalls of working with text.
Several layers are created through pitch shifting, and a
rhythmic loop seems to have been created in the lower frequency.
The repetition of this loop can be identified, but because it lacks
clear attacks, it does not provide a strong rhythmic statement.
The main layer of forward text has continual variation and
renewal that has been created mainly through editingvarying
repetitive material through changes in relationships.
There is enough variation in density to maintain overall
interest, although at times it gets a little thick. This variation in
density is important, since it allows for longer gestural shape,
which keeping everything constant would not.
The work ends well, suddenly coming into focus on the
original text.
I would have liked to hear different kinds of processing to
separate the layers more. Im not advocating using every type of
processing; however, even with the pitch shifting, the voice still has
a limited frequency range. If the composer had tried to high pass
filter the text and add reverb, for example, a new and
distinguishable layer could have been created.

TO DO THIS WEEK
You should be finishing up the musique concrte assignment.
181

Lab Eight

Take your gestures and longer sound objects and think of

ways of putting them together in a compositional sense. Recognize
that not all of your sounds may have a place within the final work!
This is an unfortunate, but necessary, part of moving from sound
designer to composer.
Consider some of the ideas presented in this lab, together with
the examples of previous student works, in formulating a
compositional strategy.

182