Aller au contenu principal

Case Study — Acousmatic & Fixed-Media Concert

In brief
  • A fixed work diffused live over a loudspeaker orchestra — playback becomes performance.
  • The composed space is already there; diffusion re-projects it into the hall.
  • Two philosophies: projecting by level (Acousmonium) or by filtered spectrum (Gmebaphone).
  • Rehearse in the actual hall; capture a diffusion score that is repeatable yet live.

The brief​

A composer presents an evening of fixed acousmatic works in concert — mostly stereo pieces, a couple in 5.1 or 8-channel — before a seated audience of about 250. The venue is a general-purpose hall rigged for the evening as a loudspeaker orchestra: an Acousmonium, dozens of loudspeakers of deliberately different sizes and characters, placed across the front, along the side walls, at the back, raised on stands and slung overhead, some near the audience and some far. There is no click track and no sequencer running the room; the pieces exist complete, "fixed" on their support, and are diffused live from a mixing position in the middle of the audience by a performer who has rehearsed in this specific room. The goal is a genuine performance of space — repeatable in its broad plan but alive and shaped in the moment — in which the same file heard on headphones becomes a different, room-sized experience.

This is the practice where space has been treated as compositional material the longest, and it repays study even if you never diffuse a note yourself, because every other case in this part borrows something from it. The artistic background is in Electroacoustic & Acousmatic Music; the vocabulary of diffusion logics is in Diffusion Logics for the Audience.

Creative problematics and reflections​

  • The composed space is already there — you are interpreting it, not building it. For a well-made fixed work, distance, width, depth and gesture are already composed into the material itself; the piece is spatial before it reaches the hall. Diffusion does not manufacture a space the work never had — it re-projects the work's existing composed space into the particular listening space of tonight's room. Denis Smalley's distinction between the composed space (built into the image) and the listening space (the real room) is the whole job description: diffusion is the act of mapping one onto the other, and of managing whether the composed image sits contained within the hall or transcends it. If you approach a spatially flat file hoping to "add space" at the desk, you have nothing to interpret.

  • Diffusion is performance, not playback. The diffuser takes a work of two, six or eight channels and spreads it live across an orchestra of many more loudspeakers — pushing a swell outward to the far and high speakers so the sound seems to leave the hall, pulling an intimate passage down to a close frontal pair, letting a gesture travel the side walls. These moves are played by hand, shaped to this hall and this moment. Two diffusers give two valid, different performances of the identical file, exactly as two conductors give two readings of one score. This interpretive status is argued directly by Jonty Harrison ("Sound, Space, Sculpture," 1998) and Annette Vande Gorne ("L'interprétation spatiale," 2002), both discussed in Diffusion Logics.

  • A serious diffusion is scored, not improvised from scratch. Expression lives in prepared, rehearsed structure with room for live inflection — not in winging it. The diffuser rehearses in the hall and writes a diffusion score: which passages go where, where the orchestra opens to full width, where to hold back to a single pair, where the risky moves are. The performance then breathes within that plan. This is the same principle as the cued spatial states in the theatre case, only played by hand rather than fired on "GO."

  • Two philosophies of the loudspeaker orchestra are both on the table. You can project a work by level — distributing amplitude across a heterogeneous array of full-range loudspeakers, the model François Bayle established with the Acousmonium at the GRM in 1974. Or you can project by filtered spectrum — splitting the signal into frequency bands and routing each to loudspeakers specialised for that register, the model Christian Clozier developed with the Gmebaphone at Bourges in 1973 (later the Cybernéphone, reaching seventy-six diffusion channels by its 2005 version). The choice is genuinely expressive: level-based diffusion sculpts where energy sits in the room; spectrum-based diffusion sculpts the colour and grain of the space itself. The full lineage — pupitre d'espace (1951) → Gmebaphone / Acousmonium (1973–74) → BEAST → today's tablet control and timeline automation — is one continuous idea.

  • Restraint and the shape of an evening. An acousmatic concert is a long-form dramaturgy of space: the diffuser paces contrast across a whole programme, not just within one piece. Reserve the fullest deployment of the orchestra for the moments that earn it, so that opening the room to its widest reads as an event rather than a default.

The cases within the case: what format is the piece?​

The source format changes the diffusion job completely, and a real programme mixes them:

  • Mono pieces. The purest diffusion: a single channel spread by hand across the whole orchestra, the performer drawing the spatial image live. Nothing is baked in, so everything is interpretation.
  • Stereo pieces. The staple. A composed L/R image is expanded across the array — front pairs carrying the image, wider and distant speakers extending it — without destroying the internal balance the composer set. Stereo is an encoding to be decoded, not two channels to route.
  • Multichannel / octophonic pieces. Composed for a specific ring (often 8 channels) — the hall's array must be mapped to the composed layout, and the diffuser has less freedom because much of the space is fixed in the file.
  • Ambisonic / scene-based tape pieces. The work is a sound field that must be decoded to whatever array is rigged — closest to modern object/scene rendering, and it tours to different orchestras cleanly because the decode adapts.

The festival reality​

Most acousmatic concerts are shared. A festival runs many composers over a few days on one rig, which adds hard logistics: each piece may want the orchestra tuned differently; sound-check time per composer is short; the diffusion desk must recall a different setup and level map per work; and someone must diffuse pieces they did not write, from the composer's notes. Recallable per-piece configurations stop being a luxury and become the thing that makes the evening possible.

Technical problematics and reflections​

  • Many discrete outputs, driven expressively in real time. Dozens of independent loudspeaker feeds must respond to a performer's hands live, with low, constant latency and no zipper noise on fast gestures. The control surface — a physical desk of faders, or a tablet — has to make sweeping a source across twenty outputs feel like one musical motion, not twenty separate fader rides.

  • Room-specific projection tuned in the actual hall. The same file is diffused completely differently in a dry black-box and a resonant church; the reverberation and geometry of the room are part of the instrument (see The Sound Field & the Room and reverberation). Setup, level-matching and rehearsal have to happen in the venue, not on studio monitors — the composed and listening spaces only meet in the room.

  • The heterogeneous array is an instrument to be tuned, not flattened. The loudspeakers differ on purpose — small and bright, large and warm, near and far — and their differences are expressive resources. But their placement, orientation and relative timing still demand time-alignment and careful gain-staging so the ear localises the way the diffuser intends and the precedence effect works for you rather than against you. Level-matching a mixed array is part of tuning the instrument, not a fault to erase.

  • Repeatable and live at once. The performance wants recallable structure — so the rehearsed diffusion score holds from one night to the next — and live override, so it can respond to the room, the audience and the moment. Automation and hands, not one or the other.

  • Multichannel and stereo works side by side. A programme mixes stereo pieces with a few multichannel ones; the system has to fold both into the same loudspeaker orchestra coherently, treating a decoded stereo field as a projectable spatial object rather than two channels to hard-patch.

Solutions — the general approach​

Treat delivery as live diffusion, not playback: a diffusion surface — a desk of faders, or a tablet — driving many outputs, rehearsed in the hall and captured as a diffusion score of states and moves that the performer then plays through. Choose the projection philosophy that suits the material: level-based across a full-range orchestra (Acousmonium-style), filtered-spectrum across specialised loudspeakers (Gmebaphone-style), or a hybrid. Tune the array in the venue — time-align it, level-match it, walk the seats — before the doors open. Admit stereo works as already-spatial fields to be projected whole, not as channels to scatter. The lineage from the pupitre d'espace to tablet control and timeline automation is one idea carried forward: diffusion is an interpretive act adapted to the room, and modern tools generalise the diffuser's desk without replacing the practice. This sits within the broader live-sound toolkit, turned to a fixed-media, sit-down-concert end, and overlaps the concert side of education & research programmes where this repertoire most often lives.

Solutions — with RIPL​

  • The diffuser's desk, generalised. RIPL continues the diffusion lineage in two directions at once: live control from a tablet during the concert, and precise, recallable automation of every spatial parameter on a timeline. The rehearsed diffusion score can therefore be played live by hand, composed in advance as automation, or both — with the performer's hands overriding the automation in the moment, exactly as an interpreter departs from and returns to a plan.

  • A fixed work as a live-diffusable scene. The stereo and multichannel pieces enter RIPL's unified source model, in which a decoded stereo field is a first-class source alongside mono objects and multichannel stems. The whole programme is projected across the loudspeaker orchestra as one coherent scene rather than hand-patched channel by channel — so a stereo work and an 8-channel work are diffused with the same gestures over the same array.

  • Render to the orchestra actually rigged tonight. Because authoring is separated from rendering, the evening's specific, heterogeneous array is simply the current render target. The same diffusion — the same scene, the same score — recalls to a different hall and a different set of loudspeakers next week without re-authoring, which is what lets a diffused programme tour.

  • Fast hand-gestures kept clean across many outputs. RIPL spends its expensive interpolation on the loud, moving foreground gesture the audience is actually tracking, and economises elsewhere, so a hand-played swell stays artefact-free even when it sweeps across the full orchestra of outputs.

  • Both projection philosophies in one engine. Level-based projection across a full-range array and register-specialised routing coexist in the same scene, so a diffuser can shape where energy sits and what colour the space has from one surface rather than two separate systems.

Pitfalls and checklist​

Checklist
  • Did you rehearse the diffusion in the actual hall, on the actual array — not on studio monitors or in the abstract?
  • Is each piece already spatial, so diffusion has a composed space to interpret, or are you trying to add space that was never written in?
  • Is there a written diffusion score — states and moves — that a second performance, or a second performer, could follow?
  • Do fast hand gestures stay clean across every output, with no zipper noise or lag?
  • Is the heterogeneous array tuned as an instrument — time-aligned, level-matched, its differences used on purpose — rather than flattened into uniformity?
  • Does the shape of the whole evening pace its use of the orchestra, so the widest deployments land as events?

See also​

In the technical guide:

In artistic practice:

→ Next: Sound-Art Installation