TUTORIAL for the HANDBOOK FOR ACOUSTIC ECOLOGY


SOUNDSCAPE
COMPOSITION


The soundscape composition is a form of electroacoustic music, developed at Simon Fraser University and elsewhere, characterized by the presence of recognizable environmental sounds and contexts, the purpose being to invoke the listener's experience, associations, memories and imagination related to the soundscape.

It grew naturally out of the pedagogical intent of the World Soundscape Project (WSP) to foster soundscape awareness. At first, the simple exercise of 'framing' environmental sound by taking it out of context, where often it is ignored, and directing the listener's attention to it in a publication, broadcast or public presentation, meant that the compositional technique involved was minimal, involving only selection, transparent editing, and unobtrusive cross-fading.

This 'neutral' use of the material established one end of the continuum occupied by soundscape compositions, namely those that are the closest to the original environment, or what might be called phonography or "found compositions". Other works use transformations of environmental sounds and here the full range of analog and digital studio techniques comes into play, with an inevitable increase in the level of abstraction. However, it is typical that the resulting sounds are abstracted from the original, that is, by retaining an aural or psychological relationship, and do not become entirely abstract.

However, the intent of transformation is always to reveal a deeper level of signification inherent within the sound and to invoke the listener's semantic associations without obliterating the sound's recognizability.
 
We will present these categories as follows, with compositional examples selected from WSP members including Hildegard Westerkamp who has specialized in this genre, along with those by composers who have been published by Cambridge Street Records. However, we encourage you to listen to works by the wide range of composers and environmental sound artists working in the area worldwide.

A) A continuum of approaches

B) Fixed spatial perspective

C) Moving spatial perspective

D) Abstracted spatial perspective

E) Detailed analysis of specific soundscape compositions


home

A. A continuum of approaches. In a rapidly evolving field that can be loosely grouped together as soundscape composition, it can be expected that there is a wide range of approaches. One way (and not the only one) to get an overview of this range is to place it on the following continuum where the approach mentioned above, phonography (by analogy to photography) is placed in the middle.

The phonographic approach is rooted in Field Recording where the intent is to document a particular place and its soundscape, or at least certain aspects of it. There is minimal transformation, and any editing or mixing is done transparently, meaning that these manipulations do not attract particular attention.

On the other hand, there are subjective choices being made at every stage of the work – which are quite clearly demonstrated in the Field Recording module with many of their dimensions – and the field recordist never regards the material as "objective" or "neutral". In that sense, the final result can be regarded as a composition that can be listened to as if it were music, that is, with the same attention one can give to music (assuming it's not just background sound).

Given this concept of phonography, we can place it in the middle of this continuum:

Sonification/Audification <----------------> Phonography <----------------> Virtual Soundscapes
      mapping the real                              found sound                            abstracted sound

Sonification refers to a range of approaches that map real-world data onto sound parameters, sometimes termed auditory display. Closely related is audification that translates the data directly into sound, i.e. as a waveform, though there are usually tricky issues of the time scale being at audio rates, or needing to be transposed in some way to accommodate human hearing.

Applications of sonification range from artistic projects driven by scientific data in installations, sculptures, fixed media works and concert pieces, such as the EcoSono projects by Matthew Burtner, to those where artistically informed design strategies are used to communicate scientific data to the public. Today many such projects are driven by environmental concerns, such as the work of Andrea Polli.

This range of approaches can be regarded as extending from “science in the service of art” to its reversal as “art in the service of science”. This distinction adds some clarity to the more general term “environmental art” where the question arises as to whether the environment merely serves the artist (as material or context), or whether the art can actually benefit the environment and public perception in some way.

Phonographic approaches, as in the middle of the continuum, are often based in environmental bioacoustics or the practice of sonic ethnography with its social and cultural focus. One of its main proponents, John Drever, raises provocative questions about representation (of what, how and for whom?) which are typical of contemporary ethnographic practice – sonic tourism vs. a more involved approach – that we raised in the Field Recording module. In other words, are we creating just another illusion of “fidelity”?

For instance, Gregg Wagstaff’s TESE (Touring Exhibition of Sound Environments) field projects have incorporated community involvement in recording and assembling a phonographic style documentary of their community. Other sound artists working in this vein spend a lot of time doing research and actually living in a community while they are field recording. They may also present their own audio representation back to the community for approval.

The right side of the continuum, from phonography to what we’re generalizing as “virtual soundscapes” with increasing abstractedness, is what we will cover in this module. What we can only refer to is the contemporary use of multi-channel speaker rigs (often in multiples of eight channels in a surround-sound configuration) that can provide a sense of complete immersion in a virtual soundscape, including height. However, many of the stereo examples we present here may suggest that direction.

To summarize the general characteristics of soundscape composition, here is an often quoted list:
  • Listener recognizability of the source material is maintained

  • Listener's knowledge of the environmental and psychological context is invoked

  • Composer's knowledge of the environmental and psychological context influences the shape of the composition at every level

  • The work enhances our understanding of the world and its influence carries over into everyday perceptual habits
Once you start working with sound recordings and apply some of the transformation techniques as covered in the Tutorial, you realize that the recognizability of the sound object can quickly be lost, or at least obscured. In many cases, that doesn’t matter because the compositional intent is not about the sound object in any semantic sense, but rather it is about the sound as sound for its own sake.

The style of composition that results from this approach is called acousmatic, which refers to sound where the source is unseen, but to give it less of a definition of absence, the sound itself is regarded as sufficient to carry meaning. The formal practice of acousmatic music puts an emphasis on the spectromorphology of sound, literally “sound shapes”, as manifest through the timbre and textures of sound, discussed in the second Vibration module

Therefore, the acousmatic approach moves sound towards an abstract dimension where contextual references can be minimized. Soundscape composition, on the other hand, wants to integrate the sound source and the listener’s contextual knowledge into the composition, and hence some degree of recognizability is maintained, as outlined in the first two points above. Listeners also may react by saying they “participate” more while listening because of the recognizability of the sounds they hear and what they evoke.

The third point about the composer’s knowledge influencing the composition at every level is actually more challenging in some ways. It’s certainly not the way instrumental music is taught in the Western classical tradition, where a similar degree of abstraction (as in acousmatic music) is assumed and preferred. In fact, music with too much real-world reference is termed “program music” and is generally given less importance.

Soundscape composition, on the other hand, intends to bring the complexity of the real-world (including the inner world of memory, affect and symbolism) to bear on the compositional result. Probably the best way to describe this subtle process will be through the sound examples in the rest of this module.

To put it even more simply: the soundscape composition is about something in the real world, and listeners are invited to participate more in listening.

The final point above about changing our perception and understanding of the world can be traced to the idealism of the young composer researchers working in the World Soundscape Project in the 1970s, but the current concerns for environmental issues have certainly re-kindled some of that optimism and desire for making a contribution.

Whether or not soundscape composition (or any other environmental art form, for that matter) will have any lasting effects remains to be seen, but once you engage with it as a recordist and sound designer, it is inevitable your experience will carry over into everyday life. Your ears will be tuned differently.

Moving beyond soundscape composition. The notion of soundscape as how a sonic environment is perceived and understood is generally associated with a concept of place and the creation of acoustic space. A more general umbrella term could be context-based composition where the approach as outlined above opens up the references to any real-world context and utilizes the entire range of approaches outlined above. The intent is to be more inclusive and to think about how we can use sound effectively to address any other issue – essentially to argue with and through sound.


Index

B. Fixed spatial perspective. There are various structural types found in soundscape compositions that derive from soundwalks, sound monitoring, and field recording practice. The first of these we will consider is characterized by a fixed spatial perspective, similar to the exercise of sitting in one place and monitoring all of the sounds one hears – a useful exercise on any occasion, and one is often surprised by what actually happens.

This basic perspective emphasizes the flow of time, or else it can be a discrete series of fixed perspectives possibly linked by smooth edits or cross-fades. Media experience seems to result in listeners being quite tolerant of simple transitions such as this without being disoriented by the replacement of one soundscape by another.

Variants on this format include time compression (where the time flow is foreshortened by leaving out parts of it, not a compression of the sound itself), as well as those involving narrative or oral history accounts that do not involve a soundwalk.

Again, the foreshortening of time seems quite natural, not only because of media practice, but because human memory does something quite similar. Recordists notice this when they listen to a long take of events they've recorded, and realize that those events now seem to actually “take longer” than the way one remembers them. We probably just remember the highlights, so to speak. Memory is clearly not correlated well with clock time.

Since a soundscape is defined and treated as the perceived sonic environment, then time compression and foreshortening techniques may actually make the edited recording seem more realistic, at least within certain limits.

Layering in stereo or 8 channels. We heard this first example in another module as a demonstration of “soundscape fusion”, where four stereo recordings, all done around Vancouver harbour, were mixed together as a single fixed-perspective document in stereo. Although there might be logical inconsistencies, most listeners would accept this as a plausible phonographic representation of the harbour, something that could have happened this way.

But of course it didn’t. In the complete recording, done for the 1996 Soundscape Vancouver double CD, the intent was to produce a 4-minute recording of the harbour that could be compared with a 1973 version on the other CD, and the result was achieved by mixing four tracks together.

However, if we had to forego any mixing or editing, that would constrain us to a few minutes before and after noon when the O Canada horn sounds daily. We would likely miss other typical events that didn’t occur within this restricted time frame, and hence our representation of the harbour would be less inclusive.

Vancouver harbour mix
from Soundscape Vancouver 1996, CSR-2CD 9701

Reminder: if you want to follow along with the enlarged spectrogram, use a Command click on the image after you begin the audio to open it in a new tab


Click to enlarge

Since we clearly had to choose which events to include, and which not, the question arose as to “how many seaplanes?”. The context here is that seaplane noise is an issue for local residents living near the harbour. As well, the harbour has the beautiful backdrop of the coastal mountain range which is often snow-capped. The physical environment is “picture postcard perfect” and often used to attract tourists.

So, maybe if we were contracted by the local Tourism Office, they might not want us to include any seaplanes, and no doubt most listeners wouldn’t notice or object to their absence. On the other hand, if we wanted to “make a point” about seaplane noise, we could add a huge number in order for listeners to get the message. So how do we decide how to proceed?

A simple solution if we want to be representative is to do a soundwalk in the area first and make a count of the seaplanes we hear. This is also a good idea for discovering what else to incorporate into the mix that would be typical. As it turned out, three seaplanes were added to the four-minute mix, as well as number of boathorns and ending with the O Canada horn.

A similar approach was taken in my Pendlerdrøm work from 1996, commissioned by a Danish group on the theme of commuting, pendler being a commuter and drøm being a dream. I was given one hour of excellent quality recordings taken from the busy Central Station environment, including the local commuter train that the recordist rode on to get there. The commission was for a stereo piece, but I had a digital 8-track Tascam playback machine which could be mixed to stereo or 8-channel, so both versions became possible.

In the opening excerpt presented here, it was easy to simulate the immersive environment of the station by simply taking four original stereo tracks and multi-tracking them, as shown below in the first production page. For the first 40” there are just two stereo tracks, before they are joined by two more, so the density builds up. These tracks were from sequential recordings, but together they fuse easily and suggest a situation that was admittedly slightly busier, though not atypically, than the actual station at the time.



Production score, beginning of Pendlerdrøm (1996), Copenhagen train station sound files

All of the tracks are kept separate in the 8-channel format, except for tracks 3&4 which, being mainly the public address announcements, could logically be coming from multiple speakers. Normally for 8-channel work, it works best to put stereo pairs in adjacent channels (ignore the unusual numbering scheme in the diagram that was in use at the time for our speakers). For the stereo mixdown, speakers on the left go to the left channel of the mix, and speakers on the right go to the right channel, and middle speakers are panned to the middle, in case you’re interested. We will return to the piece in the next section for an example of moving perspective.

Opening of Pendlerdrøm, simulating the busy train station in Copenhagen,
from Islands, CSR-CD 0101



Click to enlarge

The next example from another 8-channel work Island (2000) takes this approach a step further by multi-tracking and mixing realistic and transformed materials. In fact, for the 8-track version, there are 8 tracks of short soundfiles of individual waves on a beach, and 8 tracks of transformations of them.

Here is the opening production score for the realistic sounds, that arranges 5 different individual waves around the space (N, S, E & W), plus two more of wave ambience as a background. This pattern of waves repeats after 1 minute (which you’ll probably only notice now that it has been pointed out). The only processing is that there is a slight exaggeration of the crest of the wave breaking clockwise to the right which has been expanded by 90° to make it seem broader.



Production score, beginning of Island (2000), realistic sound files

Next we have the 8 transformed tracks, the first four being high-pass filtered so that just the very brightest component of these small wavelets is heard. These sounds are also resonated with a Karplus-Strong resonator with strong feedback. Tracks 3&4 are transposed downwards by a fifth. The other four tracks are low-pass filtered with the same resonator, but at a low resonant pitch with maximum feedback to create a drone. This drone at first seems out of place in a seaside soundscape – the intent being to add an air of mystery to this imaginary island – but in fact the same waves are triggering the rise and fall of the feedback creating the drone.


Production score, beginning of Island (2000), transformed sound files

Opening of Island (2000)
from Islands, CSR-CD 0101

Click to enlarge

Found sound with or without time compression. The first example is an unedited recording of an individual event – New Year’s Eve in Vancouver harbour, 1980-81 – which can easily be listened to as a composition in its own right. All of the boats in the harbour begin sounding their horns as midnight approaches, at their own pace and rhythm, and the “performance” lasts up to about 15 minutes. The reverberation off the mountains on the other side of the harbour helps to blend all of the horns which can be heard in different directions and at different distances.

The initial “attack” of the event, after a tentative start, is a large ferry docked near the recordist, after which many other boats join into a “steady state”, followed by a very long “decay” as the participants drift away. This excerpt is just the first few minutes of the entire sonic event.

New Year's Eve, Vancouver harbour, 1980-81,
from Soundscape Vancouver 1996, CSR-2CD 9701



Click to enlarge

The WSP's 24-hour “Summer Solstice” recordings in June 1974, were done on the rural grounds of Westminster Abbey near Mission, BC, where the birds and frogs around a pond formed an ecological micro-environment. An edited version of these recordings consisting of about 2 minutes per original hour was broadcast on CBC Radio in stereo as part of the radio series Soundscapes of Canada in October 1974.

The time compression for each “hour” was achieved with transparent editing, with a short verbal announcement of the time being the only commentary. Therefore, over the course of the hour-long program, a listener could experience the soundscape of an entire day at this site, something that would be physically impossible for an individual to do.

Summer Solstice recording 3 to 8 am, June 1974
from Soundscapes of Canada



Spectrogram of the period just before and after dawn (3:30 am)
showing frequency niches for frogs and birds (click to enlarge)


Narrative, poetry and oral history. These different types of verbal practices can be integrated into a fixed spatial perspective and provide their own rhythmic pulse.

We begin with Hildegard Westerkamp’s 1979 extension of her Whisper Study that incorporates her reading a poem by Norbert Ruebsaat “When there is no sound”, written in response to the earlier version of her piece. The poem is matched to the sound of icicles and footsteps in snow that originally appeared on Hildegard’s radio program Soundwalking on Vancouver Co-op Radio. Each element complements the other perfectly.

Excerpt of Hildegard Westerkamp's Whisper Study (1979)
with a poem by Norbert Ruebsaat
from SFU-40, CSR-CD 0501

Click to enlarge

My 8-channel work Prospero’s Voyage (2004) imagines Prospero about to leave his island, at the end of Shakespeare’s last play The Tempest. The actor Christopher Gaze intones Prospero’s last speech of the play as he addresses the audience, beseeching them to “set me free”.

The soundscape is that of a storm (in the play, Prospero has been able to command the elements) which at the end of the speech breaks into a heavy downpour that pounds the roof and is intended to sound like applause. However, note that the actor’s voice has been placed into the reverberant space of an actual theatre whereas the soundscape is clearly outdoors. Does that inconsistency actually work?

Opening of Prospero's Voyage (2004)
with actor Christopher Gaze
from Spirit Journies, CSR-CD 1401


Click to enlarge

Bruce Davis’ Bells of Percé was part of the Soundscape of Canada broadcast in 1974 and is based on recordings he (and Peter Huse) made in Percé, Québec, of the local churchbells along with an oral history interview with the parish priest.

This very poetic opening sequence filters the bells into a cloud of partials through which we hear brief excerpts of the priest’s stories about the bells, their names, and legends associated with the region. The technique of foreshadowing the main body of the piece with brief glimpses of the main narrative sets the mood for the piece and suggests some of the history, imagery and symbolism that is to come.

Opening of Bruce Davis' Bells of Percé (1974)
from Soundscapes of Canada

Click to enlarge

Transitions between fixed perspectives. As mentioned above, transitions in the context of an otherwise fixed spatial perspective do not have to be “logical” in either the time frame or how one has actually moved from one soundscape to another. However, they shouldn’t sound too awkward or abrupt either, unless a violent contrast is somehow desirable.

In Hildegard Westerkamp’s Talking Rain (1997), the opening section includes several rainy ambiences, all of which have a different density and character. However, she avoids the more usual technique of the smooth cross-fade to go from one to the other. For one thing, it is easy to habituate to the sound of rain and a cross-fade might only suggest a change in its character, not a change in location.

Her solution, used twice, at first seems irrational because the transition is signalled by a brief sound of a car passing on a rainy pavement – a very urban type of sound (whereas the other sounds of rain appear to be in a forest). However, each transition works at the immediate level of drawing attention to the change and making one more aware of the gentler texture and lesser density of the ambience being introduced. On a larger level of the entire piece, she is later going to make a sudden transition to a noisy downtown rain-soaked street in the middle of the piece, so these transitions foreshadow that contrast.


Click to enlarge either image

Two transitions from Hildegard Westerkamp's Talking Rain (1997),
Earsay CD


Westerkamp uses another type of foreshadowing to open her work Beneath the Forest Floor (1992) where individual sounds and gestures are introduced individually over a mysterious and obviously transformed “thumping” sound, thereby creating a suspenseful sense of anticipation while evoking a West Coast rainforest. The unifying thumping sound keeps the perspective fixed, but the brief encounters with other events suggest transitions in time or memory.

Opening of Hildegard Westerkamp's Beneath the Forest Floor (1992)

Click to enlarge


Index

C. Moving spatial perspective. A moving spatial perspective involves a smoothly connected space/time flow, in other words, a journey. The reference model is of course the soundwalk, although it can be interpreted much more broadly than that.

For instance, the sense of movement may be embodied in the recordings being used, or it can be simulated by various processing techniques, some of which are derived from the acoustics involved. For instance, reverberation is a key indicator of the nature of an acoustic space, and the ratio of the direct signal to the reverberant portion is a strong cue for distance perception, as described in the Sound-Environment Interaction module, and the studio demonstrations in the Reverberation module.

As discussed above, a simple cross-fade between recordings often suffices to suggest movement, even if it is unrealistic in terms of the time frame involved or the absence of aural cues such as footsteps or vehicular movement. Panning in the stereo plane can also be effective, although the “phantom image” that this creates is not very stable based on one’s distance from a stereo pair of loudspeakers.

Binaural recording builds in both external localization and a recognition of the recordist’s own movement. However, it is not always clear whether the listener will identify with the movement they sense, or whether they feel they are just observing someone else.

On the more general level of a moving perspective is the concept and experience of a journey. Throughout history, myths have often portrayed both mundane and heroic forms of the journey, usually with a level of symbolism as to what it means in human terms. Likewise, the literary form of the novel usually involves an evolution of its characters, psychologically as well as physically. Journeys, for instance, are often metaphors for a coming-of-age story, for instance, or the search for some truth or goal.

So, with this degree of richness in the semantic possibilities, what can be done purely with sound? Given its disembodied quality, sound has the advantage of suggesting both a physical movement in an acoustic space, and a mental or internal type of movement through memory and the imagination. The latter is usually suggested by a transformation of realistic sounds, and not just the old film cliché of the image getting blurred and the sound becoming heavily reverberated!

Here we will explore a variety of approaches in the electroacoustic domain.

Classical cross-fade and reverb. The first departure from the phonographic style of documentary approach by the WSP, from edited field recordings towards a more compositional approach, occurred with the preparation of a track for the Vancouver Soundscape called “Entrance to the Harbour”, published in 1973. The idea was to re-create in sound a journey by boat from the outer limits of Burrard Inlet, identified by the great diaphone (i.e. foghorn) at Pt. Atkinson, a soundmark that dated back to the early 20th century, past other foghorns at Prospect Point, Calamity Point and Brockton Point, under Lions Gate bridge, into the inner harbour. At the time, this trip could actually be made on a ferry, since discontinued, and was familiar to many Vancouverites.

Vancouver soundmarks, 1973

However, there were at least two problems. The actual trip took at least a half hour, and when the recordists rode on the ferry, they only got boat noise. However, they also had recorded the various foghorns separately as part of the documentation of sound signals in the area, and could easily record a typical docking procedure, including passengers meeting friends and picking up their luggage in a very small room with a squeaky door. Using these elements and some wave ambience, they attempted to simulate the trip in a 7’ mix using the simplest of cross-fade techniques and changing the levels of the various horns so that they appeared to approach and recede in the distance.


Spectrogram of part one
Entrance to the Harbour, Vancouver Soundscape 1973
CSR-2CD 9701

Spectrogram of part two
(click either image to enlarge)

By today’s standards, the mix is quite crude with illogical elements (the waves are clearly on the shore, even as we imagine being on a boat, and the transition to the docking sequence is fairly abrupt). Since the studio didn’t have a proper mixer at the time, only rudimentary level changes were possible. However, the foreshortening of the trip to seven minutes and the illusion of approaching and passing the foghorns, as well as the bridge and its traffic noise, were strong enough cues to make it seem realistic.

The track can appear to be a phonographic style documentary despite its flaws, but closer examination also shows a compositional structure based on the journey format that has a strong character. The beginning with the reverberation of the foghorn off the mountains and across the water suggests a wide open space and the natural geography of the area. Human sounds don’t appear until the docking sequence, and indicate that we are now entering the city itself. At the very end, the space we are in (small and acoustically bright) is the very opposite of where we started, suggesting a narrowing of the acoustic space from very large and open, to enclosed and domestic.

Using a parallel circuit. The parallel circuit, as described here, involves splitting a signal into multiple versions, each of which can be processed separately and mixed back together. A simple example of this would be a cross-fade from the original sound to a processed version. In my work The Blind Man (1979) I cross-faded the original WSP recording of the bells of Salzburg Cathedral – a very rich mix of seven bells – with a filtered version of them that essentially removes the attacks so we hear just the resonant partials. A description of this analog process and the French studio where it was done is presented here.

Text fragments from the poem by Norbert Ruebsaat were added to the bell mix, as heard next. In particular, the line “already it has come and is leaving again” – intended to describe the wind – is reflected by the bells which have rung, but then without their attacks seem to have left, leaving only their memory and resonances. Mirroring a text with an appropriate sound transformation can be quite satisfying, and help listeners understand the text on multiple levels.

Bell and text sequence from The Blind Man (1979)
on Digital Soundscapes, CSR-CD 8701

Click to enlarge

Layering the part and the whole. In La Sera di Benevento (1999), the soundscape begins in the local train station in Benevento, Italy, with a train arriving. A smooth transition in the train sounds which are increasingly resonated takes us into a kind of daydream with a fountain and afternoon cicadas. Can you spot the first transformed sounds hint, there are two.

Opening of La Sera di Benevento (1999)
on Islands, CSR-CD 0101

Click to enlarge

The first transformed sound in the above example is actually the squeaky wheel from the train at 0:29 which is looped and mixed with the train
a kind of repetition that is entirely plausible with trains, even though this squeak actually happened only once. The second sound is the whistle at 0:39 (the same one heard first at 0:12), but this time it is granulated, then stretched, which is where most listeners will identify a transformation. Listen to the opening again to hear how that worked.

The squeaky wheel is resonated, similar to the example above from Island, on pitches that are consonant with the horn, so that by 1’ the resonant drone takes over. There is a brief lull at 1:30 as the cicadas are introduced, and then we’re back to the train horn and drone. An Italian poem is introduced later. The intent is to lead the listener seamless into a daydream on a hot afternoon in southern Italy.

In fact, the repetitive rhythms of trains passing (called ostinati in music) are very tempting to process, since a simple loop taken from them is easy to edit out and repeat in synch with the original recording. A slightly more complex version of this parallel processing (by multi-tracking) was used in Pendlerdrøm, introduced above, with much larger trains in the Copenhagen station.

In fact there were two such loops used simultaneously, as shown below as the Source for tracks 3&4 and 5&6. Each loop is resonated at two different pitches (for the left and right channels), and gradually the feedback is raised until it becomes a resonant drone. Note that the percussive rhythm in the second loop keeps its rhythm going.

Tracks 1&2 use a filtered version of the high frequency brake sound that is time-stretched and resonated as well. The mix of all three stereo tracks creates a smooth daydream effect, but also notice that tracks 7&8 have a untransformed engine sound that keeps us grounded in the realistic soundscape already created.

First passing train loop

Second percussive train loop


Production score of first daydream in Pendlerdrøm (1996): click to enlarge
First daydream sequence in Pendlerdrøm (1996)
from Islands, CSR-CD 0101


Click to enlarge

The second daydream section in the piece, once we’ve settled down on the smaller commuter train, is handled similarly, with two stereo tracks being resonated, including a percussive wheel drone, and the sound of brakes an octave lower, resonated and stretched. The transition into the daydream is faster (I guess our commuter is really tired at the end of the day) and a random collage of short bits of sound from earlier in the piece (mainly voices and signals) come back in short loops that are repeated and resonated, a little like flashbacks or “earworms” that go through one’s mind during relaxation.

So, the final challenge is how to get out of this daydream. In this example, we hear the end of the drone sequence where even the resonated soundbytes are fading into the overall texture, and then as the train slows down we hear a station being announced, and there is a huge percussive chord as our commuter suddenly wakes up with a start.

In fact, this sound is derived from a simple door slam on the train itself that was heard earlier from another train where passengers were getting off, but it has been significantly enhanced with a resonator and a 5:1 granulated stretch factor. We then hear the door signal as it opens, our commuter gets off and walks down a wooden staircase into the outdoors and the piece ends with what sounds like a cigarette or pipe being lit.

Train door slam (original source)

Production score for the ending of Pendlerdrøm (1996): click to enlarge
Ending of Pendlerdrøm (1996)
from Islands, CSR-CD 0101


Click to enlarge

Because the original sound material was recorded around 6pm, the scenario of the commuter going from Central Station to home seemed a natural choice. However, for reasons only known to the Danish recordist, he actually recorded it in the other direction, going from the suburbs to downtown, so the sequences from his recordings had to be re-ordered to go in the other direction, and so far no one has noticed or objected.

The final example in this section is from Anne Holmes’ Wood on Wood on Water (1978), a quadraphonic piece with a coastal soundscape. The journey in this case is not a physical one, but is a bit opposite to the daydream in Pendlerdrøm, namely a transition into a high attention state where all of the background noises fade away.

The featured sound is a series of simple hits on a beached log, which is quite resonant on its own, but then a virtuosic percussion performance emerges (created with tape delays, although it could be mistaken for an actual percussion ensemble). The dynamic sounds move into the foreground and are so engrossing, we’re not likely to hear that the tide is gradually coming in (a common experience at a tidal beach).

A beached log improv from Anne Holmes’ Wood on Wood on Water (1978)
from SFU-40, CSR-CD 0501


Click to enlarge

In the section following this excerpt, even the seagull cries disappear, until finally at the end of the concert, there is a sudden surge of the tidal waves as we realize what has happened to the tide in the meantime. Gradually towards the end, water engulfs the log and it loses more and more of its high frequencies and resonance, until it is finally submerged, showing that wood and water are opposites in many senses. Eventually the seagulls return (also to our awareness) and the piece ends.

Other examples of a moving perspective will be presented in the final section.


Index

D. Abstracted spatial perspective. Here we will group together a variety of approaches to soundscape composition that do not fall simply into a fixed or moving perspective model. To use similar terms, we could say that they exhibit discontinuous space/time flows. However, to put it more positively, we could also say that they present their material from multiple perspectives, or in an abstracted, even symbolic manner.

However, keep in mind that in the electroacoustic era where we daily hear a mix of acoustic and electroacoustic sounds, it has become customary to hear technological sounds embedded in the everyday soundscape. Murray Schafer referred to this phenomenon as schizophonic to emphasize the split between the source and its reproduced form, and occasionally the “fit” still has a disruptive or “nervous” quality about it. However, we usually adapt quickly and normalize the presence of such embedded sounds, even as we understand that they are in the environment, but not of it.

The first example is a simple montage (i.e. edited) string of Vancouver sound signals, that was called (for better or worse) “The Music of Horns and Whistles” when it appeared on the Vancouver Soundscape 1973 vinyl recording. The fact that all of the sounds are pitched made it easy in those days to refer to them as musical, and the track itself was often the one selected by others to showcase the WSP work.

Excerpt from The Music of Horns and Whistles
from The Vancouver Soundscape (1973), CSR-2CD 9701


Click to enlarge

In the current context, there is simply a common theme to the track – all of the sounds you hear, along with an appropriate ambience, are typical Vancouver signals as a port city, the end of the transcontinental railway line, with an industrial harbour and mountains on one shore which adds a natural reverberation to many of these sounds. This excerpt is the last minute of the 3-minute track.

Next we have my Pacific Fanfare (1996), an anniversary tribute to the WSP that begins with a 30” montage of Vancouver soundmarks, the Pt. Atkinson diaphone heard earlier in this module, the 9 O’clock Gun in Stanley park, the Royal Hudson steam train whistle, the B.C. Ferry horns at Horseshoe Bay, and the Holy Rosary Cathedral bells.

Immediately following this opening, we hear many of these sounds time-stretched (as documented in the Microsound module) and resonated. Compared with the simple identification of each sound at the beginning, we now have a more reflective listening experience of hearing “inside” them as their evolution plays out. There are also new soundmarks introduced such as the light rapid transit Skytrain (with a bit of sly humour when the characteristic starting motor sound is slowed down). The piece ends with the O Canada horn, sounding at noon downtown, and in the distance, the transcontinental Canadian Pacific Railway (CPR) train whistle in the distance, a reference to its historic role in uniting the country east to west.

Pacific Fanfare (1996)
from The Vancouver Soundscape (1973), CSR-2CD 9701 and Islands CSR-CD 0101


Click to enlarge

Next we present two movements of Canadian composer Claude Schryer’s Vancouver Soundscape Revisited (1996) who refers to the piece as “an impressionistic portrait of the musicality and poetry of the soundscape of Vancouver”. In terms of this presentation, it is also a good example of an abstracted perspective on the soundscape of the city, based on the WSP’s recordings which were collected in the 1970s and 1990s.

The second movement of the piece, Fire, seems like a stream of consciousness in its often surprising juxtapositions of the city’s sounds. However, it is not at all random. Schryer catalogued the sounds he chose from the WSP collection in terms of spectrum, category, function, pitch and context, and in this movement he linked the sounds together according to one or more of these aspects they had in common.

He made only small modifications in terms of level or pitch to facilitate smooth transitions, one exception being the downward transposition of the foghorn at the very end. Fire being notoriously volatile, that metaphor allows quite free associations with the energy of these sounds, dynamic at the beginning, slowly subsiding to some final embers at the end.

The fourth movement is called Noise and begins and ends with a complaint about that kind of disruption, followed by a more literal series of examples.

Claude Schryer's "Fire" movement, Vancouver Soundscape Revisited (1996)
from Soundscape Vancouver 1996, CSR 2CD-9701

Click to enlarge
Claude Schryer's "Noise" movement, Vancouver Soundscape Revisited (1996)
from Soundscape Vancouver 1996, CSR 2CD-9701


A final example in this section is from the German composer Hans Ulrich Werner who has specialized in radiophonic work, including Vanscape Motion (1996), which like the Schryer work, was composed for the Soundscape Vancouver project and double CD in 1996.

In this excerpt from the piece, the spatial perspective is abstracted by reversing their more usual presentation. We hear the sequence of foreground signals used in The Music of Horns and Whistles from the 1973 Vancouver Soundscape publication, but they are in the background, and act as a kind of memory, perhaps emphasizing their historical nature. In the foreground is a sequence of reverberated waves and water splashes that place them in a different space altogether, one of intimacy and repose. This technique of having two different acoustic spaces with highly contrasting affect seems to facilitate reflective listening.

Excerpt of Hans Ulrich Werner's Vanscape Motion (1996)
from Soundscape Vancouver 1996, CSR 2CD-9701

Click to enlarge


Index

E. Detailed analysis of specific soundscape compositions. Here we will present a more detailed analysis of some specific pieces by Hildegard Westerkamp and Barry Truax. Although published on CDs in stereo, these pieces are usually performed in 8-channel format which provides an immersive effect for the listener which in some cases simulates an actual acoustic space or an abstracted version of it that creates a virtual soundscape.

Hildegard’s work Into the Labyrinth (2000) is one of several pieces composed in the context of visits she made to India which included workshops and soundwalks with people there. She describes the work as a “sonic journey into aspects of India’s culture. It occurs on the edge between dream and reality, in the same way in which many visitors, myself included, experience this country.” The piece has been published on the Earsay CD, Into India.

It clearly falls into our above category as an abstracted soundscape where highly realistic environmental sounds are combined with their transformations. She has generously documented several of these transformations which are presented in this sidebar. They follow the tradition of sound object transformation, including pitch transpositions, equalization and mixing variants into more complex sound clusters, similar to the studio demonstrations used throughout the Tutorial.

In these examples, notice how exploratory they are, and how at certain key points, a feature will emerge that obviously caught her ear, and subsequently was utilized in the next stage of the process, such as the resonance heard on one of the strikes in the stone chipping recording from Rajasthan, once it was transposed down 8 semitones and then an octave which stretched it further.

The examples also reveal some practical details about how to separate a sound, such as the bicycle bell from Delhi, recorded on a busy street, from its noisy ambience into a more useful form. It eventually becomes both a resonant drone when stretched and pitch shifted by 6 octaves, as well as a cloud of bells when different pitch shifts are combined and transposed.

The extremely high pitched “singing crickets” from Goa, as a whole and in two isolated recordings, also create rich textures when transposed to which Hildegard has added reverb that gives them a unique acoustic space, as in the sidebar.

Here we present a stereo mix of the opening of the piece which evocatively introduces the Indian soundscape in a series of overlapping vignettes that seamlessly mix the realistic sounds with the processed ones, or as she puts it, places them “on the edge between dream and reality.”

Opening of Hildegard Westerkamp’s work Into the Labyrinth (2000)
from Into India, Earsay CD

Click to enlarge

Hybrid Convolution. In other modules, we have covered the basic theory and practice of convolution via Impulse Reverberation and Auto-convolution. The special case of auto-convolution where the sound is convolved with itself leads to the question as to what will happen when we convolve two different sounds together.

Since convolution involves a multiplication of the frequencies which the two sounds have in common, our first expectation is that it will be worthwhile to pursue this hybridization only when there is significant overlap between the spectra of the two sounds. High frequencies convolved with low simply won’t produce very much, if at all.

It was with a curiosity to find out the answer to this question that I embarked on experiments in 2009 with what I’m calling hybrid convolution. Even the first results were so inspiring that within an amazingly short period of time I was able to create an 8-channel work called Chalice Well (2009) and premiere it at the Sonic Arts Research Centre’s Sonic Lab in Belfast on a 32-channel speaker rig which gave it the added height that was needed.

Previous casual experience showed that if you combine two broadband sounds together, you only get an undifferentiated wash of sound that isn’t useful for much other than an ambient track. So, why did these initial experiments succeed? First, the materials used were environmental recordings, mainly of watery sounds, but also glass breaking, bubbles in water, sharp percussive sounds of locks and doors, and some gated and transposed male phonemes. These were convolved with themselves, and two other types of sounds, a female vocal text, and four examples of sine-wave granular synthesis.

The key element to the success of the experiments was that all of these sounds were “pointillistic” in a sense, that is, with micro attacks in their textures, including a few major percussive events, such as a water splash. This characteristic gave the convolved sounds a similar, extended texture that seemed to be a hybrid version of the two component sounds.

The second element that was suggestive of a possible soundscape was that some of the watery sounds (recorded by David Monacchi in Italy) were deep in a well that had its own resonant properties. When extended, these sounds had a reverberant quality. So, if we categorize the source sounds as dry or wet (no pun intended), we get the following spatial result when they are convolved with each other:
Dry with dry produces a foreground sound
Dry with wet produces a middle ground sound
Wet with wet produces a more distant sound


There is a full documentation of all of the sounds used in the WSP Database, but since there were more than 200 hybrids produced, we can’t include them all here. Instead, we will put a selection of them in a sidebar.

A third variable that came out of the experiment and was particularly valuable for creating an 8-channel space was the discovery that file A convolved with B was not the same as file B convolved with A, as some texts might imply. After a quick consult with Tom Erbe, the creator of SoundHack, the reason became known.

Convolution will be symmetrical (A*B is the same as B*A) but only when the analysis window is rectangular, that is, no slope to the window. The analysis windows I was using were typically the Hamming variety which has slopes on either end. So, even though the results of A*B and B*A were similar in terms of their spectra and texture, their global amplitude envelopes were quite different as shown below.

This proved to be useful because more variants could be generated to fill up the 8-channel space. The differences in amplitude envelopes provided an extra spatial dimension for the sound to seem to move around the space. It is quite interesting to consider that this difference at the micro analysis level could make such a larger difference at the macro time scale – yet another instance of the nature of the microsound domain.


Water*Well

Well*Water (click to enlarge)


As to the macro compositional level of the piece, the image of a well came to mind early on because of the Italian source material, and my experience in visiting Chalice Well at Glastonbury in the south-west of England, a holy site steeped in legends and myths. Wells sometimes have a certain aura about them, as this one did, despite the fact that, being covered, there was little in the way of sound to be heard.

However, in reading about the many myths associated with the Well, the one that made an impression was the legend that underground caverns exist below it, and that this is where Joseph of Arimathea buried the chalice known as the Holy Grail, in order to protect us from evil. These caverns have never been discovered, and historians are quick to point out that many legends were in fact created by the monks in the Middle Ages to boost tourism (some things never change!). But, in the case of soundscape composition, we are not dependent on referring to actual space, and the sounds I was generating with hybrid convolution suggested an imaginary virtual soundscape, and a soundwalk type of moving perspective through it.

Here are two excerpts in stereo reduction from the piece, the first starting at the 4’ mark for a scene I call The Chamber of the Feminine. The name refers to the historical gendering of the well as feminine, and some of the source material included a female voice reading from the Song of Songs about a well (not that you’re going to understand any words but the files referred to below are identified as "AWell" and "AFount"). When convolved with other sounds, the female formants come through and seem blurred and sustained enough to be symbolic.


Production score of Chalice Well, Chamber of the Feminine

Excerpt of Chalice Well (2009)
Chamber of the Feminine
starting at 4'
from The Elements and Beyond, CSR-CD 1401

Click to enlarge

The second excerpt is from the end of that scene with the transition to the Glass Chamber, named because of the glass elements used in several hybrid convolutions of glass with other sounds, and at the end of the excerpt we are at the entrance to the underworld.


Production score of Chalice Well, Glass Chamber; click to enlarge

Excerpt of Chalice Well (2009)
Glass Chamber
starting just before 8'
from The Elements and Beyond, CSR-CD 1401

Click to enlarge

Extended complexity in hybrid convolutions.
My 2019 work Infinity Room tries to take hybrid convolution a step further towards a more abstracted perspective. The only analogy I could think of (after the materials were already developed) was the visual complexity and sense of boundless space in an “Infinity Room” installation, which has made them very popular in art galleries, but what could we imagine would be the aural equivalent?

The resulting work creates a somewhat analogous experience in an immersive space where a large number of environmental sounds are refracted (i.e. convolved) with each other and move around the listener until they trail off into the distance. Percussive, as well as musical, vocal and noisy sounds from various cultural sources, combine in seemingly endless variations as we move from mirrored room to room.

Technically, the work is based on a set of 27 acoustically rich environmental samples, each of which is convolved with all others to create 351 stereo hybrids (where the L/R channels are A*B and B*A). The samples range from impulse responses in large spaces through percussive sounds, musical and vocal excerpts, bells, thunder and strange noises.

Each of the 351 hybridized sounds is spatialized independently, often in trajectories around the listener in the multi-channel space incorporating echo delays. It is hard to imagine that any other soundscape-inspired work has embodied this much complexity.

Just to give an impression with a stereo reduction of this multi-channel work, here is an excerpt of three “rooms” (or sections) of the piece, the first being a variety of percussive sounds convolved with each other. This is followed by a section based on a clapping sound convolved with many other sounds, spatialized as shown below, followed by a short clip of a traditional Korean band convolved with various impulse responses from other spaces.


Cue from the 3rd section of Infinity Room (2019) with 8 stereo tracks of convolved sounds; the numbers indicate the loudspeakers in the trajectory of each sound in the TiMax2 Soundhub spatialization software

Excerpt of Infinity Room (2019)


Click to enlarge


Index

home