On Loop: New Interactions with Feedback (English version)

On Loop:

New Interactions with Feedback

(English version)

Tom Gurin

2024

Abstract

This study explores the evolution of a system enabling live interaction with a feedback loop filtered through a solid material during a musical performance. Feedback filtered through a space is often used as an interactive or non-interactive element. However, feedback filtered through a solid material is less common and has been associated with static installations. This article aims to challenge this notion by posing and addressing the following question: how can one interact with feedback filtered through a solid material during a sound performance? To answer this question, I draw on existing works before presenting my own experiments and the resulting performance possibilities.

I. Introduction

In 2022, I began working on my first sculpture project as a music composer. Seeking to establish a link between sound and three-dimensional space, I began to observe microphones and, above all, speakers as intersections between sound and space. My subsequent attempts to blend music and sculpture led to a question: how can feedback be filtered through a solid material? The result of that project was an installation in Paris in June 2023.

To take this idea further, I have developed a new question: how can one interact with feedback, filtered through a solid material, in real time during a sound performance?

To answer this question, I will begin by exploring the existing musical context by tracing the history of the works that have influenced my research. I will briefly explain what feedback is and how it works. I will focus on the aspects of feedback systems that influence my own compositions, which will lead to two important examples of musical performances (composed by Agostino Di Scipio and Alvin Lucier) illustrating how to apply a physical feedback filter. I will then discuss four pieces that explore the possibilities of interacting with feedback filtered through space (composed by Agostino Di Scipio, Alvin Lucier, Cathy van Eck and Scott McLaughlin) before moving on to two pieces that demonstrate the possibilities for interacting with feedback filtered through solid material.

In the next section, I will explain my own position in relation to other composers working with feedback. I will present my experiments and research, aimed at finding the right solid material to filter feedback during a performance, and the right musical gesture to alter the resonance of that solid material during a performance (without touching the microphones or loudspeakers). Finally, I will present the results of my research and assess the strengths and weaknesses of my system.

II. Definitions

3. Feedback Loops

A feedback loop consists of at least three elements: a microphone, an amplifier and a loudspeaker. Together, these three elements create a sonic and electrical signal loop. Beginning with the microphone, the sound is picked up and converted into an electrical signal. The microphone transmits this signal to the amplifier, which increases (as its name suggests) its amplitude. The signal is then transferred to the loudspeaker, which in turn (re)translates the electrical signal into a sound signal. Finally, the sound is diffused into the air (or another medium) and picked up again by the microphone. As Cathy van Eck, a Belgian/Dutch author and composer working with feedback, says in her book Between Air and Electricity (2017): “If the sound is now softer than the first time, the sound will pretty soon disappear. But if the sound is emitted on a higher level that [sic] what is commonly called ‘acoustic feedback’ is generated.” In the absence of outside influence, the entire system resonates at the least resistant frequency (van Eck, 2017, pp. 55-56).

Which frequency is the least resistant? The loop itself introduces inefficiencies that can make one frequency more favorable than another. For example, each loudspeaker has its own range. In other words, there are often frequencies which, even if they enter the speaker, are absorbed rather than diffused because of the technical limitations of the speaker – the magnet, the coil or the diaphragm:

Clearly, no frequency outside of the range of the loudspeaker can be the least resistant. However, in theory, the tympanic principle proposes the possibility of an ideal membrane that has zero resistance and can therefore reproduce any frequency equally dexterously. In principle, according to van Eck, “…the membrane should disappear and no hint of it should be recognizable” (van Eck, 2017, p. 67). (In reality, it’s not always so simple.) This implies, at the other end of the spectrum, an object whose resistance is so strong that only one frequency is allowed: the tuning fork (van Eck, 2017, p. 60). Of course, almost every object lies between these two extremes in terms of vibrational resistance. (It is precisely this principle that David Tudor explores in his Rainforest installations (1968-73), in which semi-resonant objects are used as loudspeaker membranes).

2. Filtering a Feedback Loop

Now that the feedback loop has been dissected, filters can be brought into focus. Filtering a sound signal means, simply put, amplifying or muffling frequencies, as a non-ideal membrane does. Sound filters are used in the production and broadcasting of various types of music, but in the case of feedback, the effect of filtering is exaggerated due to the innumerable passes the signal makes through it. Filters therefore play a crucial role for many composers and sound artists working with feedback.

Notably, Agostino Di Scipio — an Italian composer and sound artist working on emergence and chaos — uses the resonance of a large space as a sound filter in his project, Audible EcoSystemics (discussed in Di Scipio, 2022). The second part (n°2a), entitled ‘Feedback Study’ (2003) relies solely on feedback to generate the sound (Di Scipio, 2022). During the piece, the frequencies that the space amplifies and those that it muffles become clear. Writing about the entire project, van Eck explains that Di Scipio is “interested in unveiling the sound character of the system…” (van Eck, 2017, p. 139). On a poietic level, in “Feedback Study,” Di Scipio lets the space determine the music. In an interview with Christine Anderson, Di Scipio states: “In the Audible Ecosystemics project, and also in sound installations, it’s the room itself that acts like a filter or an array of filters. (That’s probably where these efforts of mine come closer to Alvin Lucier’s work, however different it may be” (Anderson & Di Scipio, 2005, p. 21).

Alvin Lucier uses a principle similar to that of Di Scipio — but in a slowed-down way — for I am Sitting in a Room (1969). Lucier was an American composer and sound artist who had a considerable influence on electroacoustic music in the United States. In I am Sitting in a Room, by alternating between two microphones, Lucier slows down the feedback loop so that the audience can distinguish the variations each time the sound passes through the space (i.e. through the filter). In addition, Lucier gives the system a foundation of sonic material: a short text, spoken aloud. Here is an extract:

I am sitting in a room different from the one you are in now. I am recording the sound of my speaking voice and I am going to play it back into the room again and again until the resonant frequencies of the room reinforce themselves so that any semblance of my speech, with perhaps the exception of rhythm, is destroyed. What you will hear, then, are the natural resonant frequencies of the room articulated by speech… (quoted in Collins, 1990).

As Lucier explains in the text, during the course of the piece, the resonant frequencies of the room are amplified by the system. After three quarters of an hour, this looping process produces a series of waves of sound, totally devoid of comprehensible speech, the consonants and vowels having been heavily distorted by the filter.

The two examples above, by Di Scipio and Lucier, illustrate the process of filtering feedback through space. What’s more, it is possible to interact with spatial filtering in real time during a performance. Indeed, Di Scipio claims that it is almost impossible not to interact with it. According to Di Scipio, the mere fact of having an audience alters the resonance of the room: “The way the members of the audience sit or otherwise occupy the room is not without consequences on the room acoustics….” In “Background Noise Study” (2003; part of the Audible EcoSystemics project), “their sonic presence becomes an integral part of [the piece]…” (Di Scipio, 2011, p. 106).

3. Interacting with Filtered Feedback

Without interaction, composers can only set and leave the filter to work “on its own.” According to the Merriam-Webster dictionary, an interaction is a “mutual or reciprocal action or influence” (Merriam-Webster, n.d.). In the context of this research, it is important to specify that to interact with feedback, a person (whether a performer or a member of the audience) must act to make the sound react. For example, in another piece by Di Scipio, Background Noise Study, in the Vocal Tract (2004-05; part of the Audible EcoSystemics project), he asks the performer to hold a small microphone in their mouth. Di Scipio explains that,

‘Staying still’, ‘doing nothing’, not even breathing (or breathing through the nose cavities, letting no air into the mouth); keeping mouth and lips in a specific vowel posture for some time, then smoothly changing vowel, every now and then – all this is not without sonic byproducts.… it may be listened to in terms of the willingness of the performer to face the consequences of her or his own actions (wanted or not). (Di Scipio, 2011, p. 107).

This time, instead of filtering the feedback through a room, Di Scipio uses the performer’s oral cavity as the spatial filter. This strategy not only allows but also forces direct interaction.

Lucier also explored ways of interacting with feedback. In his piece “Bird and Person Dyning” (1975), Lucier instructs the performer to wear a pair of binaural microphones in their ears. By turning their head, the performer alters the positions and orientations of the microphones in relation to the loudspeakers, manipulating the frequencies emitted in real time. Lucier again supplies the system with a sonic foundation: a Christmas tree ornament that plays birdsongs (Lucier, 2002). He writes, “The spatial relationships between the binaural microphones and the loudspeakers determine the geographical locations of the phantom birdcalls” (Lucier, 2002, p. 25).

Two other pieces must be mentioned here. The first is “Wings” (2008) by Cathy van Eck (van Eck, n.d.). In this work, van Eck begins with a feedback loop filtered through a room. The performers interact with the feedback by moving large panels around the stage, altering the spatial resonance. Comparing “Wings” to “Background Noise Study, in the Vocal Tract,” we can imagine that the panels moving around the stage play a similar acoustic role to that of the tongue moving in the performer’s mouth.

The other piece, “Surfaces of Emergence” (2013), is by Scott McLaughlin, an Irish composer, improviser, and researcher working with sound materiality and interaction. Similar in terms of interaction to “Lucier’s Bird and Person Dyning,” “Surfaces of Emergence” asks the performer to explore various frequencies by changing the orientation of the microphone in relation to the loudspeaker. However, unlike Lucier, McLaughlin uses the pickup of an electronic guitar. What’s more, he doesn’t add any additional sonic material to the system; he simply uses feedback to produce the sound. In addition, apart from moving the guitar within a filtering space, he uses the instrument itself as a filter by (initially) manually muting all of the strings except E, allowing only the natural harmonics of the E string to be amplified. According to McLaughlin, “Feedback arises as the path of least resistance in a given configuration, which opens the door for composition as a manipulation of configurations.” (McLaughlin, 2022).

McLaughlin describes two kinds of interaction with feedback. “For feedback systems,” he says, “I work with two planes which I will call the spatial and the energetic.” For him, the spatial plane concerns changes in the resonances of a space, while the energetic plane concerns the amount of amplification given to the system. He explains:

Gain alterations may also interact with sensitivities in the spatial plane if both are in play, leading to a complex set of relations to be explored in composition and performance. In the pieces discussed below, the energetic plane is rarely used compositionally (though it features strongly in the initial setup and tuning-in to the system) … (McLaughlin, 2022).

For McLaughlin, the spatial plane is more playable, interactive, and dynamic than the energetic plane, although the two are closely linked.

4. Interacting with a Feedback Loop Filtered through a Solid Material

In the pieces discussed above, sound enters the microphone through the air, not through solid materials. Although solid materials can help modify the resonance of a space, it is the air that primarily carries and filters sound. Both Van Eck and McLaughlin acknowledge that there is an inherent materiality to feedback. However, McLaughlin relies on air to transmit and filter his work. Van Eck, in Between Air and Electricity, admits that “it was really difficult to find compositions or musical works…[using] acoustic feedback through objects” (van Eck, 2017, p. 3). Responding to this lack, she wrote a piece that filters feedback through a music stand (van Eck, n.d.). However, she does not discuss it in Between Air and Electricity, and even writes that

Contrary to acoustic feedback through air, which has been used extensively by many artists and which has resulted in performances with extensive and varied movement techniques for influencing the sound, the use of acoustic feedback through solid material results primarily in stationary installations … the physical condition of an object cannot be easily modified” (van Eck, 2017, p. 114).

Although I agree that the feedback effect through solid materials is little used, I challenge the notions that it leads to “stationary” installations, and that the resonance of an object is too difficult to modify during a performance.

For example, “Nodalings” (1976) by Nicolas Collins is a performance that sometimes uses feedback filtered through a solid object. Collins is an American composer, sound artist and author, known for his electronic works and for his book Handmade Electronic Music, among other contributions. He was a composition student of Alvin Lucier at Wesleyan University in the 1970s. In “Nodalings,” the performer moves the microphone and loudspeaker either around a space or over a semi-resonant object (whose resonance is in the middle of the spectrum between the ideal resonant membrane and the resistance of a tuning fork). For the spatial version, a typical microphone is used, while for the material version, the performer uses a contact microphone. For those who prefer the material option, Collins proposes a few suggestions for filtering objects: a table, a bathtub, a body, a stone, a boat, etc. Regardless of the object chosen, Collins asserts that “the nodes and antinodes of the standing waves of all the possible resonant frequencies… combine to form a complex sonic ‘topography’” (Collins, 1976, p. 1). This piece is an example of performative interaction with feedback filtered through a solid object. A comparison could be drawn between “Nodalings,” in which the performer moves the microphone and the loudspeaker around an object or a space, and “Bird and Person Dyning” by Lucier (composed at the same university at roughly the same time) or even “Surfaces of Emergence” by McLaughlin, both of which use a parallel strategy but are limited to spatial filtering.

By using solid objects as filters, in addition to being able to play with the positions and orientations of microphones and loudspeakers, we can also modify the resonance of the system, as Di Scipio does in “Background Noise Study, in the Vocal Tract” and Cathy van Eck does in “Wings” (in both cases, the system is largely composed of air). Lesley Flanigan, in her feedback performances, modifies the resonance of loudspeaker membranes in real time. Flanigan is a composer and performer from New York. As well as playing with the distances between contact microphones and loudspeakers (sometimes by touching them together), she occasionally transforms an almost “ideal” membrane (according to the tympanic principle) into a less “ideal” one, most notably in “Speaker Synth” (2007). To achieve this, she applies a light force directly to the membrane, sometimes sliding around the material in search of different frequencies (Flanigan, n.d.). This brings us back to the anatomy of feedback and the role of the membrane.

To summarize the above definitions, feedback consists of three elements in a loop: a microphone, an amplifier and a speaker. Feedback filtering refers to changes in the amplification or muffling of frequencies, a process exaggerated by the continuous passage of the sound through the filter. In terms of interacting with feedback through air, we can change the resonance of a space, as Agostino Di Scipio and Cathy van Eck do in their works “Background Noise Study, in the Vocal Tract” and “Wings,” respectively. It is also possible to play with the placement of the microphone in relation to the speaker, as in Lucier’s “Bird and Person Dyning” and McLaughlin’s “Surfaces of Emergence.” Finally, similar possibilities for interaction exist when filtering feedback through a solid material, as demonstrated in the works of Nicolas Collins and Lesley Flanigan.

III. Experimentation

In the section above, the composers and pieces that are most important to my work were discussed. In the context of my own research, I build on their ideas to develop an adjustable-resonance feedback system for use in musical performance contexts. Inspired in particular by the work of Nicolas Collins and Lesley Flanigan, my aim is to explore the performance possibilities of acoustic feedback filtered through a solid material of variable resonance.

First, I incorporate into the system a large, strong, solid membrane, as suggested by Nicolas Collins in “Nodalings,” to create a complex “sonic topography” (to use Collins’ term). Through this material, I then combine two feedback loops (also suggested by Collins) to further complexify the resonances. However, instead of moving the microphones across the surface, I attach each to the loudspeaker of the other loop, crossing the signals and producing a rich sonic topography. Also unlike Collins, during a performance, the goal is to keep the microphones and loudspeakers still while bending the surface to modify the resonances of the whole system and, in doing so, interacting with the feedback loop.

This approach makes it possible to explore the possibilities of dynamic and physical interaction with a feedback loop, using a solid material to influence and modulate the acoustic characteristics of the system:

I arrived at this approximate schema after working on previous projects involving the construction of feedback-based sound sculptures. Building on these earlier experiments, the goal here is to improve the system’s sensitivity and, by extension, its musicality through a process of artistic research structured around the choice of filtering material and the means of interaction.

1. What is the best solid material to filter a feedback loop during a performance?

To answer this sub-question, it is essential to first consider the ideal characteristics of the filtering body. Each type of and piece of material has its own resonant frequencies. Since feedback arises in accordance with the least resistant frequency in a given system, a material that adds moderate resistance will have only a moderate impact on the sound. On the other hand, a material that is too resistant, that filters out too many frequencies (like a tuning fork), could limit the system to producing only silence or a single pitch. It is therefore essential to find a material that offers a balance of resistance and resonance. The variables to consider include:

1. Type of material (cardboard, wood, plastic, metal, etc.)
2. Dimensions (size and thickness) of the surface

These variables influence the density and flexibility of the object, directly affecting its filtering characteristics.

In my previous experiments, I primarily used cardboard, owing to its resonance, affordability, portability, and ability to be folded into a sculpture. However, cardboard has a major drawback for performance uses: its high degree of flexibility. Not only does cardboard lack the internal rigidity necessary to create a complex sonic topography, but it also loses its shape over time, which further diminishes its filtering capabilites.

Given the impossibility of testing all types of materials and all possible surface sizes, I selected five surfaces to evaluate:

• Plastic: 30cm x 40cm x 15mm (A)
• Wood:
  • 200cm x 100cm x 10mm (B)
  • 120 cm x 90 cm x 15mm (C)
  • 140 cm x 20 cm x 15 mm (D)
• Metal:
  • 120 cm x 70 cm x 2mm (E)

To test them, I performed (without an audience) “Nodalings” on each surface, including the instruction to return “to observe the effects of time and weather on the [sonic] topography” (Collins, 1976, p. 3). I used the same speakers for each test but, unfortunately, had to replace the amplifier during testing for board B following an electrical failure within the device.

Observations for each surface:

• A: Very rigid. Easy to perform with, but there are only three points of resonance and two or three partials coming out. Sensitive to temperature and humidity.
• B: Wide and flexible. Perhaps too flexible or too thin. Leads to a vast sonic topography with lots of multiphonics but, to be able to play with it, the amplification level has to be very low and therefore almost inaudible from a distance. At higher volumes the filtering effect is negligible.
• C: Fairly rigid, with a few knots in the wood that create interesting surprises (sudden explosions in the treble). Sensitive to temperature and humidity.
• D: Comes from a broken piano. The wood is smooth, resonant and fairly uniform. There are only a few points of resonance. Not very playable because of its linear shape. Sensitive to temperature and humidity.
• E: Too flexible. Barely supports the weight of the speakers. Difficult to find resonant points or to change the frequency.

Result: Board B was the most dynamic and best suited to musical performance. To my surprise, even the plastic board was susceptible to changes in temperature and humidity.

2. What is the best way to change the resonance of the solid material during a performance while leaving the microphones and speakers stationary?

This second sub-question plays an important role in the construction of the entire system. How and where should force be applied by the performer? Apart from changing the temperature or humidity of the room, or adding or removing material from the surface, it is necessary to apply additional force (i.e. increased resistance) to alter the resonance of a solid object. So, to obtain a sufficiently wide dynamic range, it is essential to be able to apply a considerable force while precisely controlling the pressure both spatially and temporally. In addition, the musical gesture must be sufficiently natural for the performer to sustain it during the performance.

My first attempts asked the performer to apply force with the hand:

To be effective, this action requires either A) an extremely pliable material (and therefore not very resistant or effective as a filter), such as cardboard, combined with a low level of amplification, or C) a force that is difficult to maintain for a long period.

After a number of sessions with various performers and collaborators, the most practical solution identified is to place the whole system on the floor, allowing the performer to walk or otherwise move on top of it. In this way, it is possible to apply maximum force precisely, intuitively, and without excessive fatigue, while interacting with a highly amplified sound:

Small pieces of styrofoam are placed under the corners of the surface, raising it by 1 to 2 cm off of the ground. This prevents the board from being muffled by the floor. Other objects (such as weights) can also be used for this, but styrofoam is more stable.

3. Results

Below is a link to a video of an improvisation using the system and gesture described above:

During this improvisation, I adjust the amplification several times to find the right level. This is necessary to repeat for each performer according to their weight, since the level of amplification is linked to the sensitivity of the system (as McLaughlin points out concerning the energetic and spatial planes). At the start of the improvisation, the gain is too low, leading to insufficient energy compared to the resistance and, therefore, a lack of responsiveness in the system. On the other hand, if the amplification is too high, the energy becomes too strong in relation to the resistance, leading to one or two dominant frequencies. To quote Collins in “Nodalings,” “The feedback should be tuned just above threshold level so that it remains ‘sensitive’” (Collins, 1976, p. 2).

Below are photos of another performer interacting with the system:

IV. Conclusion

While interacting with feedback filtered through a solid material presents practical challenges, this system allows for dynamic performances.

Let us return to the initial question: how can one interact with feedback, filtered through a solid material, in a sound performance? This artistic research demonstrates that by filtering feedback through a reasonably bendable wooden board measuring about one meter by one meter, a performer can change the resonance of a physical filter in real time by moving over it using their own weight and, therefore, interacting with the loop. In addition, for the system to be sufficiently sensitive :

• The loudspeakers should be placed at highly resonant points,
• Each microphone should be attached to the loudspeaker of the other loop,
• The amplification level must be carefully tuned, and
• The board should be suspended (by pieces of styrofoam) 1 or 2 cm above the ground.

There are several limitations to using this system. In particular, the resonance points can change very quickly depending on the weather. What’s more, the level of amplification must be re-adjusted for each performer, as it depends on their body weight.

A number of questions remain for future research. For example, numerous material possibilities have yet to be explored. Many more surfaces in other materials and sizes should be evaluated. It may be that a thicker metal board would be more effective, and less sensitive to the environment, than a wooden one. If so, this would open the door to an interactive public installation. To do this, one would also need to find a solution for adjusting the amplification (as noted above). In addition, an area not explored in this study is the composition of notated scores, which would require either a more reproducible playing system, a flexible notation system, or both. Collaboration with other media presents another possibility for performance. This research already creates a link between sound and movement. I have also begun exploring the possibility of mixing the system with an interactive video based on the sound. This is achieved by sending the sound signal to Max/Jitter via an audio interface. In this way, I return to the sonic representation of three-dimensional forms.

Although performing with feedback filtered through a solid material requires special attention, the system developed through this research offers a particularly rich and dynamic form of interaction. The performer not only hears the frequencies, but also feels them, adding their own body to the materiality of the loop.

 

Bibliography

• Anderson, C., & Di Scipio, A. (Autumn 2005). Dynamic Networks of Sonic Interactions:
• An Interview with Agostino Di Scipio. Computer Music Journal, 29(3), 11-28. https://www.jstor.org/stable/3681669.
• Collins, N. (August 1976). Nodalings [Musical Score]. https://www.nicolascollins.com/texts/nodalingsscore.pdf.
• Collins, N. (1990). CD liner notes for I am Sitting in a Room. Lovely Music. https://www.nicolascollins.com/texts/SittingInARoom.pdf.
• Di Scipio, A. ​​(June 2011). Listening to Yourself through the Otherself: On Background Noise Study and other works. Organized Sound 16(2), 97-108. http://dx.doi.org/10.1017/S1355771811000033.
• Di Scipio, A. (January 2022). A Relational Ontology of Feedback: Situated sound-making in the Audible EcoSystemics project. ECHO 3. doi.org/10.47041/ECHO.3.
• Flanigan, L. (n.d.). Speaker Synth. Lesleyflanigan.com. Retrieved 25 June, 2024 at https://lesleyflanigan.com/#Project/Speaker_Synth.
• Merriam-Webster. (n.d.). Interaction. In Merriam-Webster.com. Retrieved August 11, 2024 at https://www.merriam-webster.com/dictionary/interaction.
• Lucier, A. (2002). My Affairs with Feedback. Resonance Magazine 9(2), 24–25. https://knut.klingt.org/docs/resonance_9_2_feedback.pdf.
• McLaughlin, S. (January 2022). Feedback and materiality on the spatial and energetic planes. ECHO 3. doi.org/10.47041/ECHO.3.
• Tudor, D. (n.d.). Rainforest. Davidtudor.org. Retrieved 25 June, 2024 at https://davidtudor.org/Works/rainforest.html.
• van Eck, C. (2017). Between Air and Electricity: Microphones and Loudspeakers as Musical Instruments. Bloomsbury Academic.
• van Eck, C. (n.d.). Wings. Cathyvanck.net. Retrieved 25 June, 2024 at https://www.cathyvaneck.net/wings/.
• van Eck, C. (n.d.). Music Stands. Cathyvanck.net. Retrieved 25 June, 2024 at https://www.cathyvaneck.net/musicstands/.