Spirit of Rain

Spirit of Rain was released on October 25th, 2024, but there was a string of single releases that led up to the main album release. The process(es) that led to this album served as the first major methodological testing ground in this research. Unlike III, which documents an arrival, Spirit of Rain is largely a manifestation of deliberate experimentation with identity fragmentation and technological mediation.

Track List

1.Silk Threads(04:25)

2.Itzama(03:03)

3.Moon, Stars(06:05)

4.Melody Machine(07:33)

5.This Track, This Lyric(03:36)

6.General Idea(05:47)

7.Dream(03:34)

Text-Disruption Process

The lyrics on Spirit of Rain emerged from a text-disruption process that started with this project. Initially, this involved non-AI systems – factor oracle and Markov chain generators1 – that produced output based on statistical patterns in the input text. When ChatGPT became widely available in early 2023, the process shifted to use AI instead, but the underlying method remained consistent: I fed the system my own text and received output that was related, but jumbled and therefore presenting me with semantics coming from a different angle than my usual one.

This was not mainly about generating finished text, much more about creating friction and a need to change. As I wrote about the album in 2023: "Much of what is presented here is written either to create friction against other forms of expression, or written as a result of that same friction". The machine – whether Markov chain or large language model – functioned as an aid for finding other paths in a strange digital dialogue with oneself.

The process typically involved multiple generation passes, with the output then adjusted, changed, read, and given meaning through (my) human curation. I maintained a discipline of finalizing attempts rather than preserving endless variants – a method I have applied many times before, especially across my visual work as a graphical designer. What I have learned from that side of my profession is that UNDO often is not a helpful function, in fact it can work against you. Keeping the option of UNDO at hand usually means you do not finalize a step, but instead stumble along with a myriad of options, none of which have a solid foundation.

Production Approach

The album embraces a skewed, altered approach to sound design, utilizing granular synthesis, unconventional editing techniques, and extensive digital manipulation. This connects to the experiments documented from 2022 onwards involving the Elektron Machinedrum, Octatrack, Prophet 10, and various processing chains, as can be seen in the Research Timeline, especially in the startup phase.

Making Dream – excerpt from production process © Jonas Sjøvaag 2024

These production techniques are a prominent part of the vocal production on the album, and also for the album as a whole, using any tool available to alter, shift and explore, in the process largely removing it from myself and detaching it, much in the same way as the disruptive method employed when writing the lyrics.

A track would usually come from playing with the text and finding a melody for it, a method that is very hard to describe and capture because what ends up seeming like a well thought out melody that has a clear place in a song, very often come from the extreme opposite: something that sounds like nothing at all, that likely will never amount to anything. I have a recording of this entire process, and will not repeat that here, but it was explored in the video placed in the Research Timeline, near the end of Deconstruction, and it's titled Making a song - GRS. It was made for a presentation on the topic I did at the Grieg Research School in the summer of 2024.

After the melody was found, the work went on, trying to refine and elaborate, to make the pieces fit together. Very much a method in itself, because working like this shows that the further you go into one identity, the more focus it requires, leaving less behind for others, as discussed in satellite essay #1 - Identity and Personality. While I was working with this music, I also started thinking of articulating this particular friction, first identifying it as something negative, something that was frustrating and that got in the way, but later seeing it as what it really was: something that in itself generated new work.

Working inside a DAW2, relying on digital tools, imposed latency3, which in the end meant that playing live would either a) rely heavily on sampled material, or b) would force me to change the sound of the output. I did a rather extensive search for plugins that would give me the desired effects while keeping the overall latency at acceptable levels, but it was not possible. On the track Dream, for example, the total number of plugins used was around 50, across 34 tracks, and that was after the vocal was rendered as new take4. The total latency was around 3500, i.e. 3,5 seconds – way too much for live playing.

So, all of these technicalities, though loved by the producer, was very much a hindrance to the musician. It extended beyond the singer-identity also; the other instruments were produced, or made-to-fit in the same way, meaning that drum machines were usually preferred as the main rhythmical foundation, reducing actual drumming to an add-on. All in all, there was a gap I could not close here. The album and the music started to exist as artifacts from a production process much more than music emerging from a musician playing; one role completely overshadowed another, and in this case that led me to finding joy in finalizing, but also a relief when I could let the entire album go and focus on something else.

The Recognizability Principle

A guiding concept throughout Spirit of Rain's development was the balance between recognizability and unfamiliarity. As in a note I found in my archive from 2022:

Recognizability gives footing for understanding, and creates a security that allows listeners to be presented with expressions they otherwise wouldn't have the patience, interest, or competence to interpret. I assume that recognizability/unfamiliarity relate as approximately inversely proportional to each other.

This principle – embedding accessible elements within otherwise difficult musical context – runs through the album's structure. Tracks like "General Idea" demonstrate the approach: semi-complex meter changes with singable melody and electronic textures alongside acoustic presence.

Why it became a methodological site

Spirit of Rain became a purely recorded artifact – no live performances planned, wanted, or even possible. This was a limitation, and also one of the reasons it became a methodological site. The production layers in this record exist only in the studio environment, and a live version would necessarily be a different piece relying on stem playback, or complete rearrangement.

This forced a distinction – developed further in satellite essay #1 - Identity and Personality – that later became productive. The distinction made was, in essence, identifying and pushing against the boundaries of one of several musical identities.

Connection to my Research

Looking back, I learned that the project Spirit of Rain was mostly about testing and exploring freely, but it also became a method that I could employ to see what an identity could survive, and whether a purely studio-based practice could constitute legitimate musical work. Several tracks connect directly to concepts developed in the satellite essays:

• "Itzama" (2) was presented in the Itzama Stalagmite installation discussed in satellite essay #2 - Machine and ties in with the concept of trust5
• "Melody Machine" (4) was an effort6 to bring musicality and live intuition back into the project
• "General Idea" (6) emerged from experiments with complex meter and collective improvisation possibilities7

The album's development also surfaced the question that would become central to the research: as I wrote in an essay draft relating to trust in 2022, "if the performer does their best, then the music8 is good, from the performer's standpoint" – an early formulation of what would become the self-gratification criterion. Nevertheless, this album was too heavily biased towards one identity of many, and the knowledge from the producer identity did not carry over in a useful way to the others. Therefore, the search, and the work, continued.