In between the posts where I write about what I did since the beginning of this project, this is a jump into present tense, more specifically of this very weekend: here, the noise voices for part C get composed, and the first voice gets recorded as well – at least three takes.
A return of Nerdville had been on the plans for Erlanger Programm rather since its very first beginning: before anything else got defined for part C, mainly because my track Confidence had been one of my fave tracks from my back catalogue already long before the Nerdville moniker even was created. So I had wanted to do something with playing effects boxes as sound sources, and to stick with the tonal arc of Erlanger Programm, had decided to bring that into context with reverb tails of repedaled piano resonances. With the piano part already in the can, it was now time to do the actual noise part.
First from a purely technical standpoint, there would not be any big steps forward from the last expansion of the setup, maybe with the exception of the MiniBrute as a processor/noise source – essentially, it was opening that drawer and connecting the bits and pieces in some fashion. The question remained what to do musically (and I didn’t want to just “jam”).
Information Theory in Counterpoint
I started with the idea to have two voices in a noise counterpoint that were also played with different setups. When reflecting the idea of the setups, I came up with a total of four ideas that I wanted to pursue further:
- using microphones to capture ambient noises in realtime (e.g. from the street below),
- processing the piano reverb tails,
- using samples of piano notes (intervals, clusters) and playing with them,
- pure no-input setup, i.e. effects as noise generators.
So two voices (more would have been too much for what I was after for this track), but I didn’t want to make my mind up which ones I wanted to use…of course, I could have written two parts, and then play them with two of the above setups each – but if e.g. I played voice #1 with setups 1 and 2 and later decided to use the recordings of setups 1 and 2, I would have been screwed.
Now there’s approaches in information theory (e.g. like fountain codes) where you can store information redundantly and it doesn’t matter which part of the information gets lost, you always can reconstruct the information from the remaining data. I wanted something like this: two polyphonic voices coded into four instrument parts so that with any two parts chosen, you get the full-blown counterpoint. Sitting down with a few sketches I had made during the last few days, I decided to work on this.
I sat during most of Saturday evening until I was too tired to continue – but not before sketching out the general concept with a pencil, to do the clean copy the next day.
Sunday late morning, I once again sat down with an A3 sheet and a bunch of coloured pens and put all ideas together. And while I was at it, I also did the detail description for the setup for the first part (the field recording one).
The next step was setting up the gear. While I normally do this in the living room on the first floor, this time I schlepped it to the floor above, simply because I needed access to windows on two opposite sides of the house for this voice. While doing this, I came to the conclusion that I would want to record this around a full hour during the early evening, to be able to capture church bells (I was already way too late for the service, obviously).
Now there was an important discovery for me yet to make: while I’m not a good instrumentalist chops-wise, I’m quite good at sight-reading, so playing a part e.g. on piano that isn’t too tricky (and that, with my skill level, is a rather simple level) sight-reading works very well. Not so for the Nerdville part. Of course, I had been wise enough to look through the different boxes (especially the programmable ones: Line6 M9 and Zoom G3) and prepare fitting setups, but still, I found that I didn’t get to perform this to my liking before the first go or so. What’s more, after three proper attempts, I wasn’t sure if I was ready.
Summarizing this weekend’s efforts for this project, it’s:
- I was successful in writing fitting noise voices for part C, as well as in defining a general approach.
- The setup (including working with a stopwatch to reference the time in the score) works.
- I need to practice making noises after a score.
- Maybe I’ll do a few more takes for this voice. But I’m confident the remaining ones will then work out quicker.
And with that, I’m leaving real-time mode for now and leave you with old thoughts for the next few posts.