The Rocks are Quiet because the Trees are Listening

the rocks are quiet because the trees are listening (2025)

(update in process) ~~~ (update in process) ~~~ (update in process) ~~~

The Rocks Are Quiet Because The Trees Are Listening is a sonic fiction project that explores tensions within analog and digital sound synthesis through the lens of demontology, or demon-ontology, as proposed by Eugene Thacker as a framework of enquiry of not only the edges of the human condition but about that which remains indifferent to us. The project speculates over a sonic realm indifferent to human audition.

It's comprised by a multichannel sound installation, sound performances and a ~22000 word booklet extending the thoughts behind the sound/theory fiction as well as technical descriptions and explainations.

Extending the idea of the 'outside' of sound, complementary visual and graphic elements are presented constructing a fictional narrative. A CD player and CD cases serving as carriers of the conceptual apparatus of the sound piece.

conceptual synopsis

For the final project of my masters in Sound Art and Experimental Informatics at the Academy of Media Arts Cologne I decided to investigate the idiosyncrasies of different forms of sound synthesis. The original spark originated through a fascination on Cybernetic music and it's implementation on analog systems specifically. Coined by Roland Kayn, this kind of music possessed for me some of the original backbones that presented an ontological counter-point to the field of Computer Music started by Max Matthews: Kayn believed that electronic processes could offer a music of flows rather than a continuation of the score-performer duality, which was common during the 60's limitation of computers.

In my practical approach I got interested in analog feedback processes in the Serge Synthesizer specifically using closed loops of waveshapers and filters. I chose these elements because these sound processing cells are the ones that generate the most tension between the digital and the analog domain:
Wave Shapers and Multipliers are a type of transfer function used in electronics to ignite upper harmonics from simple signals, thus prone to expose the limitations of representation of the audio sampling theorems. Audio Filters do the opposite from, they can be use to sculpt signals by the integration of values over time, thus effectively filtering frequencies of signals above a threshold[^2]. They generate micro-temporal delay or simply phase shifts.

During my exercises of recursion, I found out that one of the main differences of analog from digital emulations of this processes can be located within the temporal structure that is somehow reinforced through the digital sampling process done in the Digital-to-Analog-Conversion. This temporal grid, which usually happens in blocks of samples, reinforces a rather strict sequentiality that starts getting complicated when approaching the circular topology of feedback processes. Audibly, this creates a difference that, albeit tiny, might be one of the main reasons of preferring voltage circuits recursion. The analog, for its better part, understands voltage behaving continuously, in analogy to real-life phenomena, such as air-pressure influence under a microphone. Since analog feedback loops occur grid-less, the notions of previous/posterior become blurry.

For me, feedback networks not only sound fascinating but present a powerful metaphors that encompass fields such as information theory, cybernetics, DSP and electronics, but can also be touchstones for further questions regarding the modes of being, agencies and creating music.

To present an example, one of the fascinating perspectives that I found from these experiments comes from tangents with the temporality issues found within blockchain technologies. Firstly, through the solution of the Byzantine Generals problem demonstrated by S. Nakamoto in the Bitcoin, through his proof-of-work model, and secondly by the proof-of-history model as seen in the Solana blockchain. What interests me of these protocols is that they, as well as the feedback sound making experiments, offer a different kind of way of understanding temporality, a passing of time that presents alternatives to the human perception and its limitations, such as time and space, by establishing it's own internal machine-logic.

What really drives my curiosity was an open question left in my thesis. I tried to replicate analog feedback loops on the computer using the supercollider language and doing so, I found different limitations as well as possibilities. My first attempt was trying to reach the strangeness of the replication from the bottom-up, meaning, from the replication of the technical mechanisms that allow the non-linear feedback sounds. I decided to take another approach, this time using a corpus of 3 hours of recordings and training a neural network using the RAVE model. My approach was to, this time, compress my own embodied practice within a model, in order that i could create a computational representation of the sensation of my personal approach with the synthesizer. Ultimatetly, after training the model itself, my final decision was to throw text, room resonant feedback and other material and spray them in a room with a sound installation.

1. Sound Installation

pedro a. ramírez · the rocks are quiet_soundInstallation_binaural mix_2025

Coded in the computer music environment Supercollider, the installation is comprised by different cycles on which the following actions are performed based on weighted random chances:

Sub-woofer plays back a boom
Moving the Servo Motor, which is Connected to the directional speaker
Recording of the ambience of the room (Buffer(n))
Playback of the ambience of the room (Buffer(n-1))
Playback of speech sample through directional speaker
Playback of samples from a pool of improvisations on the "acrosynthetic" patch, streamed from CD playback via bluetooth. (in different versions, another audio would be played here¹).

More importantly: Every playedback sound has a chance of being either:

being processed through a frequency Shifter (to shift peaking resonances in the feedback process
being processed through the RAVE model and played back in the space. This opens the possibility of turning the room into a transfer function itself, by recursively throwing sounds over the variational autoencoder system.

Buffer System:
Since the sound installation should remain independent from maintenance from a technically savvy person, I decided to create a cascade of audio buffers from which the reading and writing functions would be implemented in a round robin fashion. This is meant to prevent positive feedback from occurring from a single circular buffer approach.

in the first round, It would record the first buffer. On the second, play back the first while recording the second and on the third, reproduce the second while playing the third, and so on. Each round would have an accumulation of room sounds and feedback signal, but still hold some room for the sounds "to breathe".

¹:

This comes from one of my attempts to "colour" such systems with meaning. A way to present a difference between the clean systemic approach of sound installations as seen in the work of Alvin Lucier or others, where the composition would be the sonority of an autonomous running system, i.e I’m sitting in a room.

By inserting a foraneous element into the system, I aim to give the piece a different structure that somehow starts finding resonances with Bateson’s relationship of the Metaphor and the mind in Cybernetics.