home kyma Research


By Roland Emile Kuit

ReVIEWING 13: International conference on Black Mountain College
UNC Asheville’s Reuter Center North Carolina, U.S.A.

Thematic Focus: Leo Amino/The Visible and the Invisible:
Submerged Histories of Abstraction

Most pictures on this page are hyperlinks to video's.

Leo Amino, “Refractional #85” (1972)


    Leo Amino background

Leo Amino is the first artist in the United States to use plastics as a principal medium for experiment. He is responsible for the innovation of cast plastics in the history of modern sculpture. Inspired in part by the Plexiglas experiments of the Russian Constructivists as well as Bauhaus sensibilities, his embrace of light and colour as primary elements of sculptural construction anticipated the work of avant-garde American artists in the 1960s, some of whom had been his students. Amino shared a resolutely anti-conformist and anti-traditionalist philosophy with the exiles and refugees of the Bauhaus. Like fellow experimentalists of his generation Josef Albers and Ad Reinhardt, Amino was initially recognized by the cooperative Artists’s Gallery, where he received his first solo exhibition in 1940. After several one-man shows in New York, Amino was invited by Albers to join the faculty of Black Mountain College in the summer of 1946, two years after the college’s integration, where he taught alongside the Albserses, Jacob Lawrence, and Walter Gropius, and informed the education of students Ruth Asawa, Kenneth Noland, and Harry Seidler, among others. The first artist in the United States to utilize plastics as a principal material, Amino pioneered this medium more than two decades before its widespread use by other American artists, making him the successor to the Plexiglas experimentations of Bauhaus and constructivist artists Naum Gabo and László Moholy-Nagy. The artist’s experiments emerged from dissatisfaction with his attempts to incorporate colour into traditional sculptural media, anticipating the concerns of minimalist artists that would not gain widespread attention until the 1960s. Amino dedicated the second half of his career exclusively to these ideas, producing a series of “refractional” compositions with light, colour, and transparency.* 


    Imagining a 3D sonic space

I was asked to create a sonic sculpture from the image of this work by Leo Amino as a canonical artwork. This flat surface represents a three-dimensional space with colour shapes. His "Refractional #85" inspired me to imagine what it would be like to look at this sculpture from all angles. We have to imagine how the shapes and colours transforming into other shapes and colours. Imagine walking around in this sculpture and touching these coloured light forms. Furthermore, We also have to imagine the sound that emanates from them. Imagine gradations of transparent sound.

    The sonic exploration for this event

In the search for transparent colour forms and colour gradients, ways must be found in the design of sounds. Noise can be both transparent and solid. It all depends on the thickness, structure, and density of the molecules. Every molecule is sound. One method of creating sound is additive synthesis. The basics of adding sine waves to create timbres.
Example of a sine wave. Sine waves are produced by an oscillator. An oscillator is basically a signal generator that produces a sinusoidal or non-sinusoidal signal with a certain frequency. A sine wave is a pure tone and has only one frequency, which is the fundamental frequency. The frequency can be expressed in units of pitch or frequency.

    The sine wave generator

This waveform is shown here with an oscilloscope: 


By adding several sine waves with different frequencies and their different amplitudes, powerful spectral sound towers can be built. The energy of eight sine waves with the same frequency. Each sine waveform is generated by a single oscillator. Taken together, they can form certain timbres. It all depends on which frequencies and their amplitude are stacked on top of each other. 

A stack of sixteen sine waves, added step by step, produces different frequency patterns as towers of timbres by changing the amplitudes and frequencies.


    Noise and spectra

White noise refers to a sound that contains all frequencies of the spectrum of audible sound equally. White noise has the same amount of energy in each frequency band. When we look at the spectrum of white noise, it appears flat or even across all frequencies because it has the same energy for all frequencies. Every single frequency is a sine wave.

Pink noise has the same energy in the same pitch ranges. Therefore, pink noise sounds more low-pass filtered and darker than white noise. The frequency spectrum is such that the power spectral density is inversely proportional to the frequency of the signal: 1/f noise.


    Creation of frequency bands

Just as Leo Amino chose his materials, We too must choose which form(s) of sound to use for this project. We have to create sound structures that can be open or clustered as material on a molecular level. :Below we find the original colours in frequency bands:


507-526 THz

570--591 nm


400–484 THz

620–750 nm


606–630 THz

476–495 nm

Noise bands are created by using white noise that passes through a comb filter. This type of filtering is created by mixing the original signal with a delayed version of the original signal. This type of mixing, adds frequencies and removes some frequencies.

The structure of the comb filter:
Noise is added to itself by delaying 16 samples, creating bands as we see in the next figure. The noise bands are formed by changing the delay:

The sample delay corresponds to the partial. A partial is a non-fundamental tone that is multiplied by a fraction, not by integers. If we increase the number of samples, there are more bands for the noise to run through:

The same principle is used in the creation of band-pass filters. With bands corresponding to colour frequencies divided into the audio spectrum.
Noise band with frequencies for yellow:


Noise band with frequencies for red:

The term red noise is used for Brownian noise. For artistic reasons I use this term here as the distribution in the original colour spectrum above.
Noise band with frequencies for blue:



    Different ways to approach Noise  

Noise froms as distinctive signs in sound. To form such identities, the methods of production must be explored. Jan Boerman, a Dutch pioneer of electronic music, invented noise by using a sinusoidal oscillator modulated by a highly random modulation signal in the frequency field of the tunable oscillator. With this type of modulation, the amplitude of the oscillator remains constant, but its frequency is changed according to the modulation signal:

The colour of the spectrum follows the increasing random modulation signal:

In the extended version we can specify further parameters to shape this randomness and introduce Brownian noise:

These parameters are a better tunable Brownian random modulation signal for the frequency of the oscillator and a white noise random modulation signal for the amplitude of the oscillator. In this example, both the frequency and the amplitude are modulated:


    Brownian noise

A white noise with a derived feedback decay as a function of a Brownian motion. Brownian motion explores the environment. A delay module can be considered as a space.
If we change the deviation of this 1-sample delay, a larger field of the modulation path is produced.


In this 1/f 2 frequency spectrum, a more pronounced modulation path can be seen. This Brownian noise is a random walk noise. Its spectral density is inversely proportional to f 2:


It is a perfect modulation signal as a modulation signal for random improvisation.

Brownian noise is used as a modulation path for a random walk signal.
In this example, the Capytalk expression randomWalk is specified in the frequency field of the oscillator.


A more comprehensive way with an accelerated improvisation:


These examples were created with only one randomly modulated oscillator generating a sine waveform. The random movement of the frequencies is specific to the generation of noise. In both structures, the frequencies can be adjusted to create a specific noise spectrum that determines the character of the sound.

    Discover more ways to create noise  

It is important to create a toolbox of options to choose from. Different types of noise have different expressions that we need to explore.
A stack of sinusoidal oscillators is the oscillator bank. The Frequencies and Amplitudes fields both expect arrays of values, as each part of the spectrum can have its own amplitude and frequency.
The number of values in the array should match the number in the Partials field. Each oscillator follows a slow, random trajectory:

A path consisting of a sequence of random steps in a mathematical space.
The sonic spectral representation of these slow eight random trajectories:


    Next level noise generation

Random impulses as noise. By accelerating these impulses as points in time-space, the noise is formed by compression:

A static noise spectrum in 3D space. The individual frequencies are scattered as static points in this box.
These points together form all kinds of timbres, depending on their position in this 3D space:

By moving these particles to different locations in space, the energy and frequency of each particle changes the colour content of the noise.
A spectacular view of beautiful possible spectral variations:

The use of an array of elements as a language in sample space:


The spectral representation of the 256 elements used to generate noise. Each element, called a TrackNumber, can be shifted in time, creating different frequencies. In this way, tuned clusters of sine waveforms can be created and modelled in sample space.

By randomly measuring this arrangement of 124 elements, spread and bundled sine waves can be generated:

The elements are distributed over 124 points in the pattern space, but are not static. The motion is generated by an image data stream that is converted to white noise as random modulation.
Connect the dots:

 Windowed particles provide a clear enunciation to shape this type of noise as a composition on a micro level.

A window is a shape that models or envelops the amplitude of a sound. By changing the shape of the amplitude, different timbres become possible.
A spectral view of small enveloped sounds:

Jittered forms that produce noise forms, tones and more. Jitter is an irregularity in the time base of a digital signal.
Sixty-four sine wave particles are windowed by a Gaussian function:

The same sound construction uses the Gaussian function for articulation. Instead of a sine wave, a wavetable with exponential noise was used.

Noise can be assembled into a rapid palette of timbres:


    Diffusing particles to colour 

The colours in this sculpture are not pure yellow, red and blue, we see colour gradients and transparent colour mixtures. Light particles can be directed to form a gradient between the boundaries of each colour, creating an x and y path in this 2D space. We can perceive microdistances as interferences between particles.
It is interesting to investigate how a colour moves between the upper and lower boundaries.

Stochastic movements between these boundaries can be constructed as a swarm, where each sound window represents a colour pixel of the spectrum of that specific colour.

 Particles with yellow colour gradients:

Particles with red colour gradients:


Particles with blue colour gradients:

For each colour, stochastic sonic trajectories can be defined that create polymorphisms in the colour space of the colour gradients.


    Spatial morphing

Windowed spectral boundaries for the colours yellow, red and blue. The morphings between the colours are created by merging the different windowed tone pixel colours.
This process projects these particles into a 3D space. Apart from the fact that the sound routes can be defined very precisely in Kyma's graphical interface, the routes can also be varied stochastically once they have been described.

In this way, the spatial aspect of the sound becomes a truly composable parameter, an independent sonic variable that shows us how important it is to morph sounds on a molecular level.


    Polymorphic for Leo Amino ** 

The sounds are produced by clustered forms of almagmatic, windowed noise. The video consists of windowed colours in yellow, red and blue.
Windowing of transparent articulated colours through a Brownian noise that leads to new random polymorphic forms. It has become a canon of Amino's art. Kyma colour study in transparency:

Leo Amino's idea of morphing abstract forms and transparent colours in this way got me thinking about how to abstract this idea even further, in a canonical way.

Polymorphic with transpared greys on sonic noise

The desaturation of colours and the shift of these transparent grey overlapping forms to sonic articulations of noise, a new step into minimalism:

Used equipment:
Symbolic Sound Pacamara:
Reworked Kyma constructions
Roland Kuit constructions
Reworked NeverengineLabs constructions

Adobe Photoshop 
Pictures of Leo Amino’s work:
“Refractional #85” ,1972
“Drawing No. 52” ,1949 



* Introduction by Dimitris Lempesis
Leo Amino-The Visible And The Invisible

** This is a part of a chapter from “NOISE: The Concept of Everything“, by Roland Kuit

Copyright © 2022 Roland Kuit


Your work is an exemplar of what is meant by “experimental music”. Your systematic, exploratory approach and careful documentation demonstrate the process of hypothesis, testing, and validation. This process is essential to “discovering/creating” new kinds of structures and procedures for the future of music. So few people understand and practice this approach to experimental music (many people conflate the mere use of technology with experimental music, but those are orthogonal!) - Carla Scaletti



Copyright © 2012 - 2023 Roland Emile Kuit. SONIC CULTURES LAB - SONIC SPACES, SONIC MANIFEST, Laboratory of Patching and SoundLab are trademarks of Roland Emile Kuit. All Rights Reserved. Unauthorized duplication and distribution of copyrighted material violates Federal Law. This website collects no personal data.