Explanations about Sounds by Swarm Intelligence:

Simulation of a swarm consisting of several sub-swarms, in which the movements of several hundred swarm members (so-called boids) have been recorded as line traces. For 4 selected boids (shown in blue), the movements are reproduced as superimposed sound sequences in stereo sound.

The movements of all boids are based on the swarm rules discovered by Craig Reynolds. Small obstacles (tiny yellow dots) partially restrict the directional decisions of the boids, resulting in uneven movement patterns. The respective positions of the 4 selected boids are highlighted with blue markings. During the video recording, the swarm rules were slightly modified so that the swarm behavior sometimes changes.

For the selected boids, the respective positions were highlighted with blue markers. These selected boids were used here as separate sound generators to generate sounds via a synthesizer. These boids can play a tonal range of 9 octaves. Individual tones have a duration between 1 (1000ms) and 16 (62ms). The tones include the 6 Solfeggio frequencies, which are said to have a healing effect in the esoteric field. The selected frequency depends on the direction taken by the respective boid.

An (extreme) philosophical interpretation of swarm behavior is contained in a section on other Swarm Art video artworks that were created using swarm intelligence.

Explanations about Sounds by Swarm Intelligence:

Wave images are converted into sound images. Different wave formations with their frequencies and amplitudes are represented in color as changing wave patterns. The waves exactly in the center of the image are acoustically implemented as a template for a synthesizer. For the generation of the video with their sound images, a self-developed Java software was used, which uses the Java Sound API and the control of a synthesizer with the sound bank contained there. The recordings are based on 3 variable virtual wave generators, whereby each wave generator is assigned its own frequencies and amplitudes.

Preset in the controlled synthesizer are the simulation of a musical instrument via the soundbank as well as a sound generator, via which the optical wave formations visible in the display are converted and mixed to a sound sequence via different measuring points in the video image. In the sound generator programmed here, the number of measuring points of the optical waves in the center of the image is preset to be mixed to a sound; the bandwidth is between 1 and 141 measuring points, where the optical position of the individual measuring points corresponds to a horizontal line in the center of the video image, where in turn each individual measuring point corresponds to exactly one pixel. A narrow bandwidth then results in shorter sound sequences, while a set larger bandwidth produces longer and more uniform sound sequences by averaging the individual measured values. The tone generated in each case from the determined wave values of the measuring points in the center of the video image is then the result of superimposing the averaged values with the received frequencies and current amplitudes of the wave patterns. Figuratively speaking, this means that you as a listener would be sitting exactly in the center of the image and there the incident waves would be converted to tones, but here without stereo effect.

The sound duration depends on the respective optical wave formation with its frequency and amplitude as well as the set bandwidth and is additionally modified by variable time patterns selected by the algorithm itself. Tones can be extended dynamically by the algorithm by doubling and tripling over the current time pattern. 

A wave metric is also preset, which can be visually distinguished from other adjustable wave metrics by its appearance, allowing different variants of tone sequences to be generated as a sequence of different wave formations. The different wave metrics are characterized by the fact that the wave propagation is not only based on conventional concentric propagation as usual for acoustic sources, but also uses, for example, hyperbolic or also other metrics, which show up optically as different basic types of wave patterns. Sometimes also inverted wave propagation directions are used, where the waves flow from outside centrally onto the wave generator, which of course does not occur in nature. However, the algorithm here can independently specify these types of virtual wave formations created by the particular wave metric.

Each wave generator emits waves, each represented by a different rgb color value (red/green/blue). The 3 rgb color values are then mixed at each pixel to form a single hue visible on the screen, so that the overlay of the 3 waves is reproduced in color exactly in that pixel. The pixel field of the whole video image then results in colorful patterns showing the superpositions of the waves and giving a hint of their original wave patterns. Each individual rgb component of the measured and superimposed individual values in the measuring points in the center of the image is passed as an independent tone/note via the sound generator for tone generation to the synthesizer based on the soundbank with the simulated instrument, according to the settings described above. This creates a separate tone for each individual rgb value, so that up to 3 different tones of an instrument can be heard simultaneously, and these 3 tones can also overlap when listening.

The positions of the 3 wave generators change independently controlled by the algorithm itself, creating the different dynamic wave patterns and resulting in the different sound sequences from the measured values determined in the center of the image in time with the image sequences. The movements of the wave generators are algorithmically generated by means of a Java function "spotmove" developed in 2002, which has already been used earlier in several video as well as interactive artworks by opartandmore (e.g. spot series). Each wave generator is visually indicated by a small dark circular area. The 3 movements of the wave generators controlled by "spotmove" can also dynamically lead to superpositions. However, a superposition of all 3 wave generators at the same time happens relatively rarely. In extreme cases, the 3 generators then form a common place for a limited time from which similar or even uniform waves can emanate. A temporary or also longer lasting superposition of in each case 2 wave generators is made by the algorithm against it more frequently. Between the positions of the 3 wave generators there are neighborhood relations, which are led by "spotmove" temporarily to similar or uniform, but also attracting or repelling or independent movements.

A 1st version of this Sound Machine was published as Java Applet "Interference" in the internet in 2008. This version could be controlled by the Internet user by mouse interactions directly in the browser, provided that Java Runtime Version (JRE) and the Java Soundbank were installed on the PC. Because of the at that time more and more increasing security risks with the use of Java applets the execution of Java applets is no longer permitted by most browser providers (exception for example Pale Moon; see also Proteopedia). As a replacement for the original Java applet "Interference" with its non-stop output of sound sequences (music non stop), only a few examples with excerpts are shown here as video, which, however, do not allow the viewer to interact with and control the sound machine as with the Java applet. In 2021 the Java-Applet "Interference" was adapted to a higher screen resolution, which can be seen in the video here on this website (approx. 800x600).