Absurd (derived from surdus, or deaf, in Latin): that which cannot be heard or is contradictory to reason.
What if we applied the principles behind conspiracy theories — absurd premises followed by logical developments — to sound art?
Alternatively, the sweet spot can be adjusted dynamically to the actual position of the listener (in video games for instance), making a correct phantom source localisation possible over the whole listening area. Massive multi-channel audio systems that apply wave field synthesis or higher-order ambisonics exhibit an extended optimal listening area instead of a sweet spot.
To be noted, sound engineers also refer to the ‘sweet spot’ of any noise-producing body that may be captured with a microphone. Every individual instrument has its own sweet spot, the perfect location to place the microphone or microphones, in order to obtain the best sound.
The idea first came up in early 2017, during a co-creation with composer Ditte Rønn-Poulsen in Brønshøj Water Tower, a place with complex acoustics due to its architectural structure: entirely made of concrete, this 34-metre-high building has an overall round shape with a central helicoidal staircase surrounded by five massive square columns, spreading to 10 and then 20 columns before joining the outer walls, on three vertical levels.
Such a structure tends to scatter the reverberation to an extreme extent. Acoustic measurements made by Eddy Bøgh Brixen, a teacher at The Danish National School of Performing Arts, during a class in 2018 show an exceptionally long reverberation tail in the low range (about 35 seconds at 40 Hz), as well as an unusually diffuse field at low frequencies.
In reality, the room reverberation is so diffuse that two people can barely have a conversation when standing five metres apart. But what really strikes the patient observer is how alive the room is, as empty as it may be: any car passing outside of it or any pigeon flying inside of it will set the room in motion, creating wild reverberations. In other words, the room does not need human presence to be activated.
This leads to the potential absence of a sweet spot: what if the world did not need us to listen to it from a so-called central point of view? Would this sign the end of acoustic anthropocentrism? Indeed, in real life (without the use of speakers) one does not need to be perfectly placed to hear a bird, or a marble rolling on the floor. Likewise, ten people in a room represent just as many reference listening points. How to join them?
Intermediate associations: recording spaces
The groundbreaking stage of this research was conducted in October 2018 at the end of a workshop with choreographer Kitt Johnson at The Danish National School of Performing Arts, after discussing her work. Her connection to the ground and to particular areas of the stage is remarkable: very much aware of the energy she loads every specific spot with as she performs, she is able to return to any of these spots at any time and resume from the energetic state that space was earlier left with. Trying to get close to that same specificity found in her work led to several unfortunate attempts, including the creation of a human shape made of speakers facing the ground, then of a sarcophagus (made of speakers, again) for the dancer, around which the audience stood. These attempts made it clear that it was not the dancer, nor an image of the dancer, that should be the focus, but rather the space. What if we recorded spaces instead, from the ground’s point of view? The first test included mapping the room with eight omnidirectional microphones (a set-up hereafter called ‘the module’) placed 10 cm above the floor, then recreating it on a one-to-one scale with eight loudspeakers.
For those familiar, the following 8-channel recordings might sound like regular binaural takes made with a dummy head[1], yet they are absolutely not. To be listenable in a stereo environment though, they have been downmixed to binaural. The use of headphones is therefore recommended.
This idea of recording spaces in which — coincidentally or not — something happens led to the concept of 'invisible choreography': a room in which a choreographic work (in the broad sense of composed movement) happens, being reproduced without its interpreter; a room that bears a trace of someone or something that has loaded it with energy, all around and eventually through its audience; a room with no preferential place to be, as curiosity might push the listener to go where the energy lies.
Recording a room does not come without issues, though. First, how to capture the entirety of an object, in our case a container of some sort, without having any distance from it? Our microphones have to be placed inside the room they are recording, with the awareness that close proximity to walls and corners will amplify reflections and therefore potentially create phase issues. All this means we will virtually create a room inside the room at the recording stage, allowing as a side-effect the possibility to be outside of the performance when stepping out of the performing cluster of speakers (this cluster more or less defining a ‘useful room’). And without proper omnidirectional speakers, which are expensive and hard to get, that question might become even more crucial.
Related to this issue, when playing back the recordings, one might need an equivalently proportioned room in order to create a proper acoustic illusion for the audience. If the diffusion room happens to be much bigger than the recorded room, for instance, then walking outside of the cluster of speakers will be like listening to a room (the ‘useful room’) inside a room (the bigger recorded space) inside a room (the even bigger diffusion space). Likewise, if the diffusion room is much smaller than the recorded room then one might hear ‘beyond’ the existing walls, which will once again break the illusion. Let’s say, at the very least, the distances to some of the walls will have to be respected from one room to the next, so the audience can use physical references.
Now to the holographic — or holophonic — effect: here we have essentially a horizontal plane, as the microphones were all placed about 10 cm from the ground. The highly spatial impression comes, for instance, from the marbles recorded as they roll on the floor, but there is no third dimension present — or at least not at this stage. Of course, one could add vertical layers of microphones but this would generate other complications. Indeed, even though the system is technically extremely simple and can be organised at will, it requires many microphones, preamplifiers, and loudspeakers plus a lot of cabling. To add a vertical plane would require two or three times as much. We are therefore facing a simple set-up, very simple math, no sweet spot, but a lot of equipment. Likewise, the generated sound files can take up a lot of space.
Following the description from the first example (silence), we should add that this project is not the first to use Rosalind Krauss’ paper within the field of sound design. Seth Kim-Cohen’s book, In the Blink of an Ear,[2] is a notable precursor there.
Without denying Kim-Cohen’s remarkable work and its pre-existence to this research, it should be mentioned that it was not a reference until very late in the process of creating this exposition, which was made without prior knowledge and towards a different goal. Still, a comparison of the two parallel developments should also serve as a healthy reminder that one can use such a system in many ways, as they widely differ (we use realistic and abstract sounds, whereas he uses noise and speech), leaving remarkable gaps in both cases (what about noise music if music exists between non-noise and non-speech? Where will we place the use of lyrics in sound poetry if everything should be either abstract, realistic, non-abstract or non-realistic?). This perhaps makes it an oversimple model, as already acknowledged by Kim-Cohen himself.[3]
Therefore, the model should be thought of as a shortcut to some possible breakthroughs, and a useful tool for analysing what already exists and what might be missing, as we keep questioning sound creation.
Example: Sleeping machine
Moreover, if we now modify the structure of the sound by means of, for instance, cutting, looping, or reversing the signal (none of which inherently change the sound’s structure), then we expand our non-realistic view: to the original choreographic sound composition we add a layer of electroacoustic composition.
Adding to that a few electronic elements to accompany the movements in a non-abstract way should confuse the audience and place them in the conditions required to experience the invisible choreography as an absurd sound composition.
The experiment: composing (with) spaces
In December 2018, during an artistic residency at Inter Arts Center, Malmö, Kitt Johnson was recorded using a mini-module made again of eight omnidirectional microphones, but this time with a scale of 0.5 to 1 at a height of one metre from the ground. The idea was to record within a smaller space: in a module such as the one described beforehand, which was about 5.5 x 3.0 metres, tiny movements would not be much represented. Therefore, aware of Kitt’s use of space, it was decided not to ask her to move much in terms of distance, but to focus on one area and load it with energy.
The inconvenience of such a small module is that, in order to recreate it, one will need a combination of eight speakers placed close to each other, meaning the audience will be unable to enter the space. This experiment in space then became coupled with an experiment in time: doubling the duration of the recordings will indeed lower one octave, which would therefore double the wavelength of all recorded sounds. This effectively restores the size of the original module, with two notable exceptions: the sound is one octave below, and the original space recorded — including the choreography inside of it — has doubled in size. The dancer’s arms would now span four metres.
New questions: the omnipresence problem
Parallel to this, another question arose linked to absurd sounds that could prove crucial in varying the sonic effects in such a set-up: can movement be abducted from space? For instance, would it be possible to make something constantly move from one side to the other, so that the usual spatial rules disintegrate in-between an audience member’s ears?
A possible point of departure is the Shepard Risset tone, an auditory illusion of a tone that continually ascends or descends in pitch, yet which ultimately seems to do neither. The problem with this as a starting point, however, is that spatialisation is either linked to amplitudes (the signal is louder to the left) or time (the signal comes earlier to the left). Unfortunately, while our ears are rather precise when it comes to pitch, they are significantly less so when it comes to amplitude or envelope. This means that all attempts to create a sound constantly moving from one side to the other resulted in blurry, phasing elements which seemed to more or less never leave their original position.
This led to the absurd idea of using the last out of the four characteristics sound engineers use to describe sound: timbre. What if we panned sounds into space depending on their frequential content? This means thousands of parts of a similar sound could move from one side to the other independently. Think bass, medium, and high frequencies moving separately, but on a way more precise scale.
Such experiments barely scratch the surface of the many possibilities when looking away from the widely accepted notion of a necessary sweet spot. These could, of course, be developed in many exciting ways: using such techniques in cinema would redefine its basic axioms in terms of reproducibility, as each member of the audience might hear something different during the screening. What would then happen to storytelling?
As a final point, removing the sweet spot entirely is not simple, as the audience to some degree must remain more or less within the limits of a cluster of loudspeakers, which perhaps makes the cluster a new, wider sweet spot. But if the sweet spot describes the room, it should also be possible to add microphones (and afterwards speakers) outside of the cluster, so we would recreate the outer walls acoustically, or even the outer world. In the end, the limit becomes physical and technical — or should we say the limit becomes our chosen stage.
Besides, what would be the point in mapping the entire world one-to-one anyway?
Storytelling is central to the associated piece of music, ‘inside an architect’s dream’, as the old water tower becomes a performing character. Here we have a huge building, which in the beginning of the piece is to the right, while someone is in front of its door, outside and to the left. A car passes, which is reflected from both points of view — the recordings are simultaneous.
The person opens the first door, then the second. The two points of view are dramatically different. When the two characters (building and man) meet they start making music. The ensuing piece was performed inside the building through six speakers placed at various heights and recorded with three stereo sets (transforming the original piece into a new one), while synthetic elements, vocals, percussions, and waterphone were performed live. The piece ends when we climb the stairs to the upper floor, 34 metres above. The room is still with us, waiting.
Analogue synths developing on floating structures are at the centre of the piece. Recorded with microphones inside the room, they are now seemingly played by the tower, as their perceived structure drastically changes due to its acoustics. We perform around the room, inside it, for it, and with it, and this reunion creates a new piece.
The results, using a rolling whiteboard and then up to a hundred marbles as examples, were stunning: people listening to the recordings would jump in the air when the marbles would cross the room in front of their very eyes — except there were no marbles. Even more, the sweet spot was lost, as one could stand wherever in the room and experience marbles rolling wherever they would roll (which enables us to ask if there ever existed anything such as a sweet spot in real life).
Is the sound generated by dancers remotely interesting as such? Doesn’t it depend mostly on the floor that supports them, as Brian Kane[4] suggests? What if a dancer jumps? Will he/she temporarily vanish? How will a dance work in a two-dimensional space, knowing that from a plane perspective up and down are one and the same thing? Also, could a choreography even be summed up by its sound? At some point in her performance Post No Bills”[5] (2015), Kitt Johnson’s fingers are moving in a spasmodic way while the rest of her body is immobile. The impression is fantastic, as it focuses the whole attention on what should seem like a detail. Using only sound, this effect would be quite an achievement unless additional accessories are connected to the body to make it more expressive. But then again, if we use such accessories, wouldn’t we risk making the body just one accessory among many, as all props become the centres of attraction?
Surprisingly, after simplifying the controls for an otherwise highly complex effect, and with the added possibility of modifying all parameters live, this became an extremely musical instrument.
Whether it be for music, theatre, cinema or videogames, sound engineers and sound designers are used to working with the notion of the so-called ‘sweet spot’: that area where all wave fronts arrive simultaneously, also called the ‘reference listening point’ as it places the listener in an ideal situation in which the sound is best heard. It is usually a small and exclusive area: out of it one loses focus on the sound image and the mix shifts out of balance.
More technically, the sweet spot describes the focal point between speakers, where an individual is fully capable of hearing the audio mix the way it was supposed to be heard. Typically, in a stereo set-up the sweet spot creates an equilateral triangle together with the left and right loudspeakers, all angles being equal to 60 degrees. In complex surround systems different methods exist to broaden its area, so that more than one listener can enjoy the sound experience as intended by the audio engineer, including the desired phantom source locations, spectral and spatial balance, or even degree of immersion.
As a concluding piece, a series of experiments were performed in January 2019 at The Culture Yard in Helsingør, Denmark, in an 800-seat auditorium. Its chairs were occupied by 120 loudspeakers, individually calibrated for each of the eight sonic parts to come — some including a sweet spot, some dissolving it, some treating each loudspeaker separately. The audience for Who's speaking? would be sitting among them, as willing subjects of a series of experiments conducted by the loudspeakers on humans.
It makes little sense to try and share the impressions left by 120 loudspeakers while sitting among them in photographs or necessarily limited stereo audio. Still, the short video below documents some of them.
A few elements that were significant to the work: to prepare audience members for the absurd sounds that would come, they were first brought on stage — the sweet spot for a massive audience applause — with the cheering loudspeakers already seated in front of them. Through sound, the loudspeakers later became, in turn, individual humans patiently waiting for the show to start, popcorn popping in a boiling pot, sonic cloud formations in the expansion of the universe, geographic rows of a naturalistic sea as its waves travelled through the room, frequential atoms of the same sea as it was performed live, conveyors of an invisible creature evolving through the audience, and, finally, emotional participants in a requiem.
LEFT SIDE: Kitt Johnson was recorded using three sound costumes created with natural fibres by costume designer Charlotte Østergaard, with the intent to amplify the sound of her movements. This short extract aims at showing the technique's potential.
RIGHT SIDE: The next step is achieved as three grand pianos (or three rooms each containing a grand piano) are recorded for a composition to be performed in a bigger space: the rooms are now used as individual musical instruments, the audience eventually navigating around them.
Let's now use this diagram to understand which kind of manipulation we are operating towards our audience, while creating an invisible choreography.
A room is recorded, and inside this room is a dancer. If the recording is reproduced as such, ideally in an equivalent room, we are then clearly working with realistic sounds — in fact as realistic as can be. Yet, performing such a recording in relative darkness and allowing the audience to feel the dancer move through them will likely make the experience non-realistic, as the dancer becomes a creature with powers unheard of.
The resulting effort, created by programmer and sound artist Xavier Descarpentries using Max/MSP, solved the problem of omnipresence, or ubiquity (how can God be everywhere simultaneously without being the same everywhere?). It became possible to move things across space without following any panning law, as their frequential content simply changed over time. Instead of a constant sound travelling continually from left to right, we’d let a simple falling sound move from speaker to speaker in a complex multichannel set-up, in which each loudspeaker is exclusively playing a certain frequency or band of frequencies. Again, no sweet spot here, but another paradigm: for the same sound, let’s say the sea, each person in the audience will hear a different result depending on where they are, as the loudspeakers closest to them will display a unique mix of frequential content.
