field + code

This is the first field + code release – two generative tracks and a bonus recording on the bandcamp download.

Each release to consist of the following:

1) A recording you can own – name your price. This one can be found here on bandcamp:

2) A patch or set of patches or other code required to run the music. This one (and likely those in the near future) is composed completely in Max, and is be available as a Max project. Here is the download link, just be warned it is a large download, as it contains all the recordings:

https://drive.google.com/drive/folders/14FDkRPV9XveZlBXu2Qh5G2Ot2MGqxvop?usp=sharing

These are tested and functioning fully on Mac, but due to some externals not being available on Windows you may see a few errors and find parts of the patch inactive.

Sharing the patches is very much part of this. I think being able to see the score running is part of what makes generative music, or music from code interesting. All I ask, is that if you use the patches please own the bandcamp version – even for free, as this way I get an idea of the interest there is in the work.

3) Discussion around the patches and processes within them. This will take place on Lines. Here is the thread for this release:

https://llllllll.co/t/field-code-01-max-patches/43234

4) Patch / process notes. Here they are for the two tracks in this release:

The field recordings for this track are taken from an evening of recording in the middle of January this year. They are from around the river Bure in Norfolk, made after a period of unusually heavy rain turned the (usually very sedate) river into a slightly faster flowing thing! One recording is taken close to the water from a bridge over the river, the other in a neighbouring field. The music around these is largely sample-based, and is constructed from a small set of Eurorack recordings. There are quite a few notes dotted around the patch as comments that help explain what is happening in the various stages of the patch. I’ll not reproduce all of that here, but will expand a bit on the process around the track.

At the core of the Max patch, there are a set of Eurorack recordings around the 4ms Spectral Multi-band Resonator (SMR) running through various filters and effects. When experimenting with the Eurorack system, I find that I often create patches that have a sound which I like, but the structure is not sequenced enough to make it a standalone track. These are often self-playing, and I’ll record them out in 10-20 minute takes to do something with at a later date. This was one of those recordings. I have then selected parts of the recording, shaped these further in a DAW and created a bank of samples from it. These samples are played back using three simple players that randomise the order of playback from one file to the next. All the files are slightly different lengths, and this helps to generate fluctuating phase relationships between the samples – sometimes they swell in sequence like a slow arpeggio, sometimes they play together. I remember at some point listening to a Brian Eno interview – where he spoke of a early installation piece he designed that used a set of (I think 6) CD players on shuffle, each with a CD of small sections of music. It struck me then as being so simple, yet effective generative technique, and it is something I’ve tried to re-create here. The two field recordings also work in the same way, and based on chance it is possible to hear all or none of these sounds at any one time.

Alongside the field recording and SMR sounds, there is also a bank of procedurally generated percussion sounds from the Intellijel Plonk. These are triggered throughout the track. They are connected to metronome with a variable clock, so there is a randomised rest time between each. At the left of the patch a set of sounds play from a non-linear sample player – that either meanders around or jumps through the sound file, with playback in both directions being possible.

Another more general approach across both tracks, is to (a) fan out the audio to work in parallel often, and (b) to add in many subtle fluctuations where possible. You can see (a) over the structure of both tracks, and it is something of an experiment to try and create soundscapes that are subtly complex and slowly shifting at all times. And in terms of (b), at many points there are gentle detunes or changes to amplitude that shift very slowly. Many of the settings for these instantiate between a range (ie cycles between 30 seconds to several minutes) so every time the patch is loaded up, these settings are different and the modulation runs at different rates.

This track has a single field recording that runs the full length of the composition. It is a dawn recording made at a friends house, I think around August 2019. I was initially a bit disappointed by the level of mic noise in this recording (it is from a borrowed Sennheiser Ambeo, decoded here to binaural) but I think it has found a place here as a low volume bed for the track. A nice thing about mixing recordings into a composition like this is that you can set the level against other instruments – so less chance someone is going to listen to a recording at the wrong level if it should naturally be quiet.

The pitched elements on this track are all tuned filters excited by noise generators. There are several flavours of noise in the patch, but the one I’ve used the most is an implementation of a velvet noise algorithm – within the patcher velvet. There, you will find an mc (multi-channel) patch with a threshold limit that only passes something like 0.00001% of the noise impulses, which then go on to ping the filters. Filter frequencies are generated by probability and these are distributed around the patch. As with the approach to track a, there are a few slightly different implementations of this noise and filter pinging combination in the track – some using fffb~ others with reson~ running through similar processing chains structured in parallel. I’ve also tried to make the tunings slightly imperfect, so you will find some pitches scaled ever so slightly to create subtle variations – as if these were an ensemble of acoustic instruments. These are sent through slowly shifting filters, some subtle distortion and a couple of instances of the vb.mi.clouds~ external (Mutable Instruments Clouds port by Volker Böhm) before the final mix stage. There is a lot of parallel processing here, and in a sense it’s very freeing to work in this way in software – where your cpu is really the only limit.

It’s a bonus recording. A very quiet one….