Thinking the Sound Space: Stage & Venue
- Space is compositional material, on a par with pitch, rhythm and timbre — not a finish added at the end.
- Every work negotiates two spaces: the composed space it imagines and the listening space it is played in.
- Map the scene: fixed vs moving sources, figure/ground, perceptible density, and gesture as phrase.
- Test every move against being material, not gadget (necessity, structure, legibility, portability).
Before you decide how to spatialise — which technique, how many speakers, what software — there is an earlier and more important question: what is space doing in your piece? This chapter is about that question. It treats space not as a delivery format but as compositional material, and offers a way of thinking about it that holds whether you are writing a fixed acousmatic work, scoring a live performance, or building an installation.
The argument leans on ideas the technical chapters explain in depth — distance and air in The Sound Field & the Room, envelopment in direct, diffuse & envelopment, reverberation in reverberation — but you do not need them to follow it. Read this first for the intention; follow the links when you want the mechanism.
Space as a parameter, not a finish
The single most useful shift in mindset is to stop hearing space as decoration and start hearing it as a dimension of the composition, on the same footing as pitch, rhythm and timbre.
This is not a new idea; it is one of the founding insights of post-war music. Karlheinz Stockhausen argued it explicitly in his 1958 essay Musik im Raum ("Music in Space"): the position and movement of sound in space is a musical parameter that can be composed, not an accident of where the loudspeakers happen to be. He put it into practice repeatedly — in Gruppen (1955–57), three orchestras surround the audience and sound is thrown between them; in Kontakte (1958–60), a rotating loudspeaker captured by four microphones encodes an actual spatial rotation into the tape. His most extreme statement of the principle is the Helikopter-Streichquartett (1992–93), in which the four players of a string quartet perform from four separate airborne helicopters, their sound and image mixed down to the audience on the ground, the rotor noise composed in as a "second instrument." The lesson underneath the spectacle is simple and portable: placing and moving sound is composing, not garnishing.
You do not need helicopters to take the point. The moment you decide that a sound belongs far away rather than close, or that a figure should cross the room rather than sit still, you are composing with space. The rest of this chapter is a vocabulary for doing that deliberately.
Two spaces, and the mapping between them
Every spatial work negotiates two different spaces, and most of the craft lives in the relationship between them.
- The composed space is the space inside the work — the scene you imagine and build: this sound is near, that one is distant and to the left, this texture wraps all the way around, that gesture rises overhead. It exists in the piece regardless of where the piece is played.
- The listening space is the real room the work is heard in — its size, its shape, its acoustics, and the loudspeaker system installed in it.
This distinction is Denis Smalley's, from his writing on acousmatic music (he draws the pair composed space / listening space in his 1991 essay "Spatial experience in electro-acoustic music," and returns to it in his 2007 article "Space-form and the acousmatic image"). French acousmatic writers, following Michel Chion, use the equivalent pair espace interne / espace externe — internal and external space. Whatever the terminology, the working reality is the same: your composed space has to survive being played inside a listening space you may not fully control.
Smalley calls the way the composed space sits inside the listening space containment versus transcendence. Played in a small room, the composed image is contained within the walls; played in a cathedral, it may seem to transcend the space entirely, or be swamped by it. The same piece is a different experience in each. Thinking in these terms early — where will this be heard, and how will my imagined space fit that real one? — is what separates a work that travels well from one that only ever worked once, in the studio, on your monitors.
Design the composed space as clearly as you can, then ask honestly how it will map onto the listening spaces the work will actually meet — a calibrated dome, a reverberant church, a flat black-box, a pair of headphones on a train. The Diffusion Logics chapter is entirely about managing that mapping.
Mapping the scene
Treat the composed space as a scene you can map. A few axes make the mapping concrete.
Fixed versus moving sources
The first decision for any sound is whether it has a place or a path. A fixed source has a location — it anchors the scene, gives the ear something stable to hold. A moving source has a trajectory — it becomes a gesture in time (more on this below). A common mistake is to make everything move; motion means nothing without stillness to move against. Decide which of your materials are the anchors and which are the travellers.
Foreground and background
A scene has depth, and depth is organised as figure against ground. Smalley, analysing a soundscape, describes how a permanent element — a river's continuous texture — becomes a "generalised ground" that anchors the whole space, while more individuated sounds read as figures against it. You compose this deliberately: what is the continuous bed that holds the space together, and what are the events that stand out from it? Distance is one of your main tools here — the direct-to-reverberant ratio and high-frequency roll-off that push a sound back are covered in distance & air and direct, diffuse & envelopment.
Perceptible density: how much can a listener follow?
There is a hard limit on how much simultaneous spatial information a listener can actually track, and it is easy to exceed. Smalley is candid about it from the listener's side: "I cannot listen to everything simultaneously and need to devote attention to each of the zones in turn, accumulating a global view, which emerges over time." Past a certain density, individual sources stop being individual — they fuse into a mass (he notes the threshold at which a texture becomes "congested"). This is not a flaw to fight; it is a compositional resource. You choose whether the listener is meant to locate discrete sources or to bathe in an undifferentiated field, and you keep the number of simultaneous, independently-moving spatial voices low when you want the former. Six things moving at once is usually four too many if you want any of them to be legible.
Gesture and trajectory as phrase
A sound's movement through space is not a special effect; it is a phrase, with the same hierarchy as any musical line. Smalley explicitly scales spatial motion from "a gesture, a phrase" up to "the stretch of a longer piece," and gives it a small grammar: a trajectory can wipe across the image (his "vectorial wipe"), and such a gesture resolves in one of a few typed ways — restoration, shift, or prolongation. Trevor Wishart, in On Sonic Art (1985/1996), thinks in the related terms of sound-images moving and transforming within an imagined landscape. The practical upshot: compose trajectories the way you compose melodies — with direction, arrival, and consequence, not as aimless drifting. A gesture that goes somewhere and lands is a phrase; a gesture that merely wanders is spatial noodling.
Depth and distance, incidentally, need not come from real movement at all. Smalley notes that the shift from far to near is largely a change in spectral resolution and level — so you can open or close perspectival space by spectral means alone, without a single positional automation move. This matters when your listening space cannot reliably reproduce position but can reproduce timbre.
Reading the venue as an instrument
The listening space is not just a container to be tolerated; at its best it is a collaborator. Some of the strongest spatial works are conceived for a specific place, treating its architecture as part of the instrument.
Iannis Xenakis is the clearest case. For the Philips Pavilion at the 1958 Brussels World's Fair — a building he designed, for a work (Varèse's Poème électronique) whose sound travelled along "routes" across some four hundred loudspeakers — the architecture and the spatialisation were a single conception. His later Polytopes (Cluny, Mycenae) took existing sites — ruins, monuments — and made them co-authors of the piece. Bernhard Leitner, working from the visual-art side, spent a career on what he called Tonraum (sound-space): building spaces you experience by moving through them, where sound has body, mass and direction and is used almost architecturally, to construct a corridor or a volume out of moving tone.
The practical question this raises for any maker is: are you writing a portable work that must adapt to many rooms, or a site-specific work that belongs to one place? The two demand opposite strategies. A portable work protects its composed space against the variability of listening spaces (see Diffusion Logics). A site-specific work does the reverse — it studies one room, its resonances and sightlines and reverberation, and builds the piece into it so tightly that it could not be moved without being rewritten. Neither is better; they are different projects, and confusing them is a frequent source of disappointment.
Reverberation as material
Reverberation deserves singling out, because artists habitually under-use it — treating it as a wash applied at the end rather than as a compositional dimension. The technical mechanism is in reverberation; the artistic point is that the character of a space is composable independently of where a sound sits in it.
This separation is exactly what a tool like IRCAM's Spat~ (developed from Jean-Marc Jot's work, then Olivier Warusfel and Thibaut Carpentier) makes explicit: source position and room character are independent parameters. You can place a sound close and dry, or far in a cathedral, or — more interestingly — close and in a cathedral, which no natural space would give you. Late reverberation is where envelopment is manufactured; early reflections are where apparent width comes from. Treating these as materials — a dry intimate voice suddenly opening into an impossible hall, a rhythmic figure that acquires and sheds its space — gives you a whole axis of expression that has nothing to do with panning. In the theatre chapter this same idea returns in a stronger form: reverberation as the construction of a fictional place that does not exist.
Material, or gadget? Five tests
The single question that separates spatial work that matters from spatial work that merely impresses is whether space is carrying the piece or decorating it. Composers who have thought hardest about this converge on a handful of practical tests — apply them before you commit a placement or a movement:
- Non-substitutability. Does this spatial move communicate something you could not communicate any other way? The electronic composer Max Cooper frames it as: what meaningful experience can I give you through spatial sound that I could not give you otherwise? If a pan could be replaced by a level change or a filter with no loss, it is probably decoration.
- Structure. Does the spatial element cover a range of perceptually different states, so that it can carry musical development — or is it just, in Natasha Barrett's words, "a setting within which the main active elements unfold"? Space earns its place when it is a carrier of structure, not a backdrop ("Spatio-musical composition strategies," 2002).
- Necessity. Kaija Saariaho's test, put plainly: "I am interested in spatialization, but under the condition that it's not applied gratuitously. It has to be necessary — in the same way that material and form must be linked together organically." Every movement should have a reason the music can name.
- Legibility. Can the audience actually perceive the gesture, from where they are sitting or standing? A spatial move too subtle to register, or readable only from the one ideal seat, is cost without benefit (this recurs as a practical constraint throughout Diffusion Logics).
- Portability. If the effect only survives on one specific rig — one array, one format, one room — that is a symptom of gadget. If the idea remains legible when you move it to a different system, it is compositional. (Marije Baalman makes this the crux of the tension between spatial composition and spatialisation technology.)
The common thread across all five: the technology serves an idea you already have; it does not replace it or supply it. That is exactly the inverse of the gadget, where the device is the point. Hold onto this when the sheer capability of a big immersive system tempts you to move sound because you can — the question is always whether the piece needs you to.
None of this forbids the spectacular. Stockhausen's helicopters and Xenakis's pavilion are spectacular and necessary — the spectacle is the idea, not a coat of paint on it. The tests distinguish spectacle that is the work from spectacle applied to the work.
From concept to score
Intentions this rich need to be captured before they meet a format. A few lightweight notations help:
- Zone maps — a plan of the venue with the composed space sketched onto it: where the anchors sit, where the audience is, which regions do what.
- Trajectory sketches — drawn paths for moving sources, annotated with timing and with the "landing" of each gesture.
- A density budget — a simple count, per section, of how many independent spatial voices are active at once, so you notice before the mix does when you have exceeded what a listener can follow.
- A cue list — for anything that unfolds in time, the spatial events in order, independent of the technique that will later realise them.
The point of scoring the intention separately from the technique is that it lets you change the technique — swap ambisonics for point-source, adapt to a different room — without losing the idea.
Tools as thinking aids: an aside on RIPL
Software does not think for you, but the shape of its controls quietly shapes what you imagine. Two of RIPL's design choices are useful as ways of thinking, independent of the product:
- A single continuous Spread/Focus control runs from a sound that is everywhere (spread uniformly across the array) to a sound that is an exact point. Because it is one continuous, automatable axis, it invites you to treat "how localised is this sound" as an expressive parameter you can play through time — tightening a diffuse pad into a pinpoint, or letting a point dissolve into the room — rather than a fixed setting chosen once.
- Every source has an aperture / directivity — a sense of how widely it radiates, and therefore how many speakers carry it. Thinking in these terms nudges you to imagine a sound as a body that radiates into the room, with an orientation and a width, rather than as a dimensionless dot to be panned.
Both are just controls; the value is that they name, and make playable, exactly the compositional decisions this chapter is about.
A worked example: spatial dramaturgy of a short piece
Suppose a two-minute piece for a modest immersive room. You have two principal materials: a low sustained drone and a bright, articulate, bell-like figure.
- Assign place and path. The drone is an anchor — fixed, low, central, a generalised ground. The bell figure is a traveller. Decide this first; it is the spine of the spatial form.
- Set the depth. Start the drone close and slightly dry so it reads as intimate ground; keep the bell figure far and reverberant at first, a small distant event against the near ground. The figure/ground relationship is now doing structural work.
- Compose one real gesture. Once — not repeatedly — let the bell figure wipe from distal to proximate: it approaches, crosses the front of the room, and lands, resolving as a shift (the drone answers by opening its reverberation, as if the room itself had grown). One legible spatial phrase is worth more than a minute of drifting.
- Respect the density budget. At the climax you might be tempted to multiply the bell into a swarm — but past a handful of voices it will fuse into a congested mass and the gesture will be lost. If you want the swarm, commit to it as texture (bathe, don't locate); if you want the figure, keep it singular.
- Check the mapping. Everything above is composed space. Before you finish, ask how it survives the listening space: in a dry black-box the opening reverberation move must be synthetic and generous; in a live church it may already be there for free — and the intimate opening drone is the part now at risk. The composed space and the listening space are always in negotiation.
That is the whole method in miniature: decide what space is doing, map the scene, compose gesture as phrase, budget the density, and negotiate the imagined space against the real one. The next chapter takes the result of this thinking and asks how to deliver it to a room full of people.
Bibliography
- Baalman, Marije A. J. "Spatial Composition Techniques and Sound Spatialisation Technologies." Organised Sound 15, no. 3 (2010): 209–218.
- Barrett, Natasha. "Spatio-musical composition strategies." Organised Sound 7, no. 3 (2002): 313–323.
- Chion, Michel. Guide des objets sonores : Pierre Schaeffer et la recherche musicale. Paris: INA / Buchet-Chastel, 1983.
- Leitner, Bernhard. Ton:Raum / Sound:Space. (Monograph on the Tonraum sound-space works.)
- Saariaho, Kaija. "A Conversation with Kaija Saariaho" (interview by Clément Mao-Takacs). Music & Literature 5 (2014).
- Smalley, Denis. "Spatial experience in electro-acoustic music." In L'espace du son II, ed. Francis Dhomont. Ohain: Éditions Musiques et Recherches, 1991.
- Smalley, Denis. "Spectromorphology: explaining sound-shapes." Organised Sound 2, no. 2 (1997): 107–126.
- Smalley, Denis. "Space-form and the acousmatic image." Organised Sound 12, no. 1 (2007): 35–58.
- Stockhausen, Karlheinz. "Musik im Raum" (1958). In Texte zur Musik, vol. 1.
- Wishart, Trevor. On Sonic Art. Rev. ed., ed. Simon Emmerson. Amsterdam: Harwood Academic, 1996.
- Xenakis, Iannis. Formalized Music: Thought and Mathematics in Composition. Rev. ed. Stuyvesant, NY: Pendragon Press, 1992.
- Jot, Jean-Marc, and Olivier Warusfel. "Spat~: A Spatial Processor for Musicians and Sound Engineers." CIARM, 1995. (On the independence of source position and room character.)
A note on sourcing: dates and attributions above reflect well-established accounts of these works and writers. A few fine bibliographic details (exact edition years, page ranges) are worth confirming against a primary source before print.