First Steps in Generative Music

Generative music is when a track is not created directly by the musician, but with the help of preconceived algorithms.

The term itself was introduced by Brian Eno in the context of his own work. On the release Discreet Music (1975) he used generative methods of music creation for the first time.

Human presence in the creation of generative music is not completely eliminated, but remains minimal. The musician acts as an operator who sets up a set of rules by which the music is created, but variation and diversity emerge through randomizers.

History

The first attempts to create an autogenerating system date back to the 18th century: in 1757, music theorist Johann Philipp Kirnberger wrote Der allezeit fertige polonoisen und menüettencomponist, a book that allowed, in game form, to generate a polonaise or minuet using two dice. The book consisted of pre-recorded short musical fragments of works and a table of numbers: with the help of cubes phrases were assembled from fragments, and after them - the whole work.

The first steps in generative music

In the early 20th century, composers such as Anton Webern, Yefim Golyshev, and Olivier Messiaen began to use serial composition techniques-when a series of unrepeated tones within a phrase are used as a system.

As soon as the first semblances of computers appeared, musicians began to use them as an instrument: in 1956, University of Illinois professors Lejaren Hiller and Leonard Isaacson wrote Illiac Suite: Quartet No. 4 for strings using the Illiac I study computer. Based on their experiments, a book Experimental music: Composition with an electronic computer was published in 1959.

Following the techno-enthusiasts, in the 1960s academic composers decided to turn to the autogeneration of material with the help of gadgets. Steve Reich began experimenting with sounds recorded on tape (It's Gonna Rain (1965), Come Out (1966) and two musicians performing the same tune Reed Phase (1966), Piano Phase (1967), Violin Phase (1967). The composition was created by the difference in tempos of the fragments recorded on tape or of the parts played. The different tempos resulted in a constant displacement of the parts relative to one another.

The same looped tape technique was used by Brian Eno in the '70s. His classic album Music for Airports used seven tape recorders. By the '90s Eno had moved on to music generation software. The album Generative Music 1, which used SSEYO Koan software, was the clearest illustration of the generative method.

The essence of the method

Generative method is most often used with modular systems, because modules have wide possibilities of modulation and randomization. Random parameter value generators, non-linear sequencers, mathematical and logical operation modules are responsible for generating patterns and phrases. Modulators allow you to combine the necessary set of elements and use almost any element of the system as a source of modulation. For those who can not or do not want to spend money on hardware modules, there is a software environment VCV Rack, where you can create a virtual modular environment, and integrate it with DAW.

But it's not just about modules. Instruments that help you create generative music can also be found in standard, classic analog and digital synthesizers. These include sequencers, arpeggiators, LFO oscillators, envelopes, Sample and Hold. Some modern models of hardware synthesizers have a Probability function, which sets the probability of triggering from 0 to 100% (example: Elektron Digitone) or a polyrhythm function, which allows you to set a different length of steps for each track of the sequencer and their constant offset (Moog Subharmonicon).

Among DAWs, the most convenient ones for generative music are Ableton combined with the Max4Live software environment and Bitwig.

How to get started

The generative approach to making music is very different from playing musical instruments.

Most importantly: the musician no longer creates or plays music himself. Instead, he has to come up with a system that plays it based on set rules. In this, the method is similar to programming.

The most illustrative example is modular synthesizers. These systems have modules for random number generation, sample & hold, noise generators, function generators, LFOs, and other devices. Due to the open architecture and free routing of the control signal within the instrument, variations can be created endlessly.

Examples of modules:

Behringer S&H / Random Voltage 1036

Behringer 150 Ring Mod/Noise/S&H/LFO

Befaco Rampage

Erica Synths Pico RND

Erica Synths Fusion Modulator

But the first steps can also be done in the DAW. For example, you can use the Chance parameter of the Ableton Live MIDI sequencer: it controls the probability of note triggering and can have values from 0 to 100%.

In the MIDI sequencer settings, you can find a "Randomize" button that allows you to give random values to a selected parameter, such as Velocity or Chance. The following standard Ableton MIDI plug-ins can be used to diversify, enliven and add unpredictability to your sequences:

Arpeggiator (for creating arpeggios, has many parameters for generating a variety of sequences);

Random (designed to generate random values);

Velocity (allows you to customize velocity, has the ability to randomize);

Scale (provides keeping within a given gamut, useful when working with randomization to avoid too chaotic sequences).

Another interesting technique is the creation of polyrhythmic patterns from several sequences of different lengths. This is exactly the method used by Steve Reich. This involves creating several MIDI tracks, setting a different length for each, prescribing the notes, and routing them to one or more instruments.

The first steps in generative music

You can achieve timbral variety by modulating through LFOs and envelopes, linking them to the right parameters. For example, you can direct an LFO with a random waveform to the starting point of a sample in a Simpler instrument, thereby changing the sound of the sample each time it is triggered.

Plugins

In addition to the standard tools, you can find many third-party plug-ins that allow you to create automated sequences:

A series of plugins from Audiomodern (Chordjam, Playbeat, Riffer, Filterstep and Gatelab);

Polyrandom for Ableton;

XO by XLN Audio;

Magenta Studio.

What to do next.

The generative approach by itself can be a full-fledged alternative to traditional composition techniques, especially when you need background royalty free music.

But it can also be used in conjunction with these techniques. Autogeneration can be used to create interesting playbacks of a live band, as the source of the base or texture of a track, or in the context of a jam - when a machine responds to a live instrument phrase with its own generated phrase.

Acquiring a wealth of experience in writing articles on trends and prospects for the development of the game industry in the world