Before examining AIYA in greater depth, it is important to give a brief background on the evolution of the Chaosflöte, the sensor-augmented flute from which AIYA draws information from. The Chaosflöte is comprised of electronic attachments affixed onto a Boehm C-Flute, a standard instrument in modern Western music and the one typically found in Western orchestras. There are two main electronic attachments on this instrument: a microphone and a second module containing three buttons, two switches, a slider, and a motion sensor . The information recorded by these electronics (audio, motion, tactile) are streamed wirelessly to a computer connected on the same WiFi network via Open Sound Control (OSC).
I created the Chaosflöte as a way for me to control live electronic elements (e.g., electronic sound processing or generation of projection visuals) during a performance. This has been particularly liberating for improvisations, where the electronic attachments served as an additional interface for spontaneous expression. For example, the motion sensor allowed me to map my natural body movements to the audiovisual expression of the improvisation, and I felt even more connected to the audiovisual electronics as a result. Being able to enact live audio processing on my amplified flute sound could enable me to envelope the entire performance space in a way that would not be possible with my human playing alone.
Chaosflöte as an extension of my body
The Chaosflöte soon began to feel like an extension of my body. On one side, the Chaosflöte served as an interface between the physical actions of myself as a human and the resulting electronically-generated sounds and visuals. On the other side, the whole system – including the electronically-generated media – began to feel as extensions of my body as well.
Over the course of five years performing with this instrument, I propose the following main attributes as being responsible for this feeling of “extendedness” between the Chaosflöte and I:
- Latency between input to the interface and expected output is reduced as much as possible (40ms is the threshold for perceived latency).
- The translation between interface input and audiovisual output is done intuitively.
- The expected output and the actual output of the audiovisual electronics are consistently aligned as closely as possible.
- The interface is sufficiently “practiced” to a point where interactions with it become more intuitive and spontaneous, as opposed to planned and calculated.
- An intuitive relationship and/or similarity can be drawn between the sonic identities of the human-generated sound and Chaosflöte-specific sound. It is worth noting that, although these are principles that can be extrapolated to other interfaces besides the Chaosflöte (e.g., AIYA or those made by others), the actual assessment of these attributes is an entirely subjective manner that will differ from user to user.
Creating a sonic identity for the Chaosflöte using loops and live-sampling
I developed a practice of improvisation with the Chaosflöte and grew particularly fond of its ability to utilize various forms of loop/buffer playback controlled by the Chaosflöte. I am aware that working with loops is nothing new, having been introduced in the 1950s, made famous by Terry Riley’s early work, “The Gift,” and subsequently appearing as a prominent device in Alvin Lucier’s I Am Sitting in a Room. However, playing with loops offers a couple of attractive qualities:
- It gives the performer the ability to fill the time and space of an improvisation, giving the performer precious moments to step back, reflect, think, and maybe even catch a breath.
- It draws an intuitive relationship and similarity to the source sound (e.g., flute, voice) and contributes to the perception of the machine as an extension of the performer’s body.
Having one slot for recorded buffers, with each recorded buffer occupying a maximum of 15 seconds, after which it is overwritten by the new incoming sound information. These buffers would have additional signal processing applied to it , so as not to sound like an exact copy of the original recorded audio.
Example A: Flipping one of the switches on the Chaosflöte to the “on” position activates the playback of a 10-second buffer. Flipping the switch off turns off the playback and allows it to resume recording.
Example B: Similar to Example A, with the exception that the player can adjust the playback of the recorded buffer using the motion sensor (e.g., mapping the x-rotation of the flute/motion sensor to the playback speed of the buffer).
Prelude to an Avenoir (2017), Chaos Ver. 4 (2018), NEXTION (2018)
Chaos Ver. 1
Having three slots for recorded buffers, with each recorded buffer occupying a maximum of 10 seconds, after which it is overwritten by the new incoming sound information.
Example A: Tapping on one of the three piezo microphones  on the Chaosflöte activates the playback of the 10-second buffer. Tapping it again turns off the playback and allows it to resume recording.
Example B: Each of the three buttons on the Chaosflöte is mapped to one buffer slot. Pressing one of the buttons on the Chaosflöte activates the playback of the 10-second buffer. Pressing it again turns off the playback and allows it to resume recording .
Selective Memory (2016)
Florida Improvisation Patches (2018, 2019)
Having five slots for recorded buffers, with each recorded buffer having a different length.
Example A: Buffer 1 = 5 seconds, Buffer 2 = 7 sec., Buffer 3 = 10 sec., Buffer 4 = 15 sec., Buffer 5 = 20 sec. The slider of the Chaosflöte determines the number of buffers playing at any given time (e.g. slider at max. left position = 0, slider at max. right position = 5).
Furthermore, it affords the flexibility to act as a counterpoint to something other than myself or become a texture while the machine takes the center stage. With the flute, it is more difficult to play chords than for example a stringed instrument; here the loop enables me greater polyphonic freedom, where I could stack repeating tones on top of each other. I can create dense, complex atmospheres layering together multiple loops. Performing with loops also changes how I approach my improvisations, in the temporal sense that I know that the material I would be playing at that moment could eventually be used as material for the future. In essence, it feels as if I am no longer improvising “alone” when I can occupy the sonic landscape with these loops.
Below are some example loop configurations I have used:
 Previous versions of the instrument also included an airflow sensor.
 e.g. pitch-shifting, exaggerated reverb, distortion effects, among others.
 A very early version of the Chaosflöte used three piezo-electric microphones instead of buttons
 In this case, a small GUI was necessary to monitor the state of the buffers (if they were turned on or off), as during the improvisation, it was difficult to keep track of their states.