Note: If you see this text you use a browser which does not support usual Web-standards. Therefore the design of Media Art Net will not display correctly. Contents are nevertheless provided. For greatest possible comfort and full functionality you should use one of the recommended browsers.

ThemesSound and ImageEmotion Machine
Granular Synthesis
Ulf Langheinrich

I am going to use some productions by Granular Synthesis [1] , in other words my colleague Kurt Hentschläger and I, to demonstrate various strategies for working with sound and images.


Granular Synthesis' work «SWEETHEART» (1996) provides an excellent illustration of some problems arising from sampling sound and images. When working with scanned and sampled material, there is inevitably a connection between sound and image within a recording. The coherence and causality that emerged in the original context of the scanned continuum are broken by cutting. This cutting is the only intervention. One particularly interesting feature here is the contrast between the causality the editing cancels out and the image and sound material that is entirely coherent in every time unit (in every ‹grain›). This is because the micro-cuts create a unique mixture in the border zone between morphing and cutting: a new synthesis that masks the cut again. This pseudo-organic situation asserts coherence. The interventions we make are actually laid open toexperience through the face portrayed, the viewer's mirror, his alter ego, as it were. By witnessing our inquisitive zooming into a few frames, the viewer's eye becomes more that of a voyeur. He becomes involved in those vehemently asserted, but groundless and meaninglessly painful conditions. These are read on to the manipulations by the ‹person› portrayed, as appropriate emotional and physical reaction. So the granular re-synthesis seems to cause the violent emotion, and triggers—now very causally—an emotional reaction in the viewer. «SWEETHEART» is the portrait of an emotion machine.

In comparison with the pictorial material, the sound material was manipulated much more extensively: the original sound was heavily compressed, so that a sensible quantity of sound is available within each 25th-of-a-second frame. The sound-picture does cite a techno-aesthetic, but is on the whole much more three-dimensional than the image tableau suggests. One essential element in the work is a seductive quality, a lack of distance in the encounter.

This also applies to the typical large installations and performances we have produced over the years, even though these have a distinctly aggressive component. This is partly due to the unusual weighting of planes in the overall sound, which is usually described as oppressively loud, even though the audience—at least in later productions—can still hold conversations. This almost absurd sound picture, almost an immersion in sound, is marked by intermodulated sub-basses, a moderate level in the audible field and chased high frequencies; it represents a physical commentary or a kind of acoustic meta-environment for the granular audio-visual stream, those flickering and twitching events that impinge via screens and loudspeakers.


«NOISEGATE» (Vienna, 1998) is the last work by Granular Synthesis devoted exclusively and extensively to the theme of «face» and «body.» This marks the end of insisting absolutely on the portrait as the negotiation location. The sound aesthetic discussed above was also devised here. At the «NOISEGATE» première at the Museum of Applied Arts MAK it was possible to experience a completely uniform sound picture in themiddle of the central space. Avoiding any directed movement provokes an attention situation that no longer enables and demands observation, but places viewers within an overall sound-picture. Events always occur in the whole field, which makes them mutations of the field and not objects in the space. Viewers watch like listeners in the dark. They can experience the full extent of the sound space and the consistency of the sound mass. The signal is felt once [on the one side] diffusing more strongly into the depths of virtual infinity, and once [on the other] taking place very compactly in the middle, or in your own head. From this central space, visitors moved to the screens in the outer rooms, where the projected material, Michael Krammer's head, was controlled interactively. But we soon abandoned this activity because the animation of the Krammer heads on the screens turned into an animation of the audience, resulting from their movement, which was not what we had intended.

We developed software to address the idea of interactive control in real time; this software drives individual computers in which sound-image grains can be accessed in the computer's RAM in real time. And in fact our new software worked wonderfully well. We could only begin to imagine this new tool's potential in the context of «NOISEGATE'S» requirements. This software was the platform for a new aesthetic. And «NOISEGATE» had clearly transformed itself into a non-interactive space dominated by dark negative images with individual harsh positive image-flashes. A kind of underground field.


The work «SINKEN» (Lille, 1999) is composed for a string orchestra and audio-visual image stream. The string parts were written by first of all defining the frequency ranges of the individual string instruments and then realizing them via individually driven AKAI samplers with the appropriate library. What can be heard are very slow glissandi over a range of one or two notes, which were later played live by a real orchestra. All the string players have to play individually, there are no groups in which musicians can get their bearings in relation to each other. This lonely playing within an amorphous sound consistency, creeping down, snail-like and slow, is a beautiful image.The video image is cloned four times. The image and its electronic sound section move even more slowly at first. At the very beginning, both are entirely still and mute, which causes a certain element of frustration to flicker up in the audience at the outset, until they acclimatize themselves to the work's slow breath. We reward them, as you see, with a beautiful shimmering sequence at the end, while the string players paddle in the lower registers.


Transition to a quasi-abstract and abstract, empty image opens up a new theme: in «SINKEN,» a real starting sample was still available, even though it could scarcely be decoded, but in later works like «FELD» (2000) or «RESET» (2001) a synaesthetic affirmation had to be set up first of all. Hence in the samples for these works, which were generated in several synthetic steps, a visual and a sound movement were brought together. These are to be thought of as a single object: every aspect of the image and the image movement has a corresponding facet in the sound, and vice versa. Stroboscopic time-light events cropped up as early as «NOISEGATE» as harsh flashes, since our new software offered that possibility, and they increasingly took over the pictorial field. Even these now emancipate themselves from torturing the alter ego electronically and become a constant compositional plane, as complex image encapsulations. Now the sub-bass touches the audience, along with shimmering light amplitudes. They are standing in vibrating sound and light and looking at emptied color fields.

This is a very good description of the situation in «RESET,» Granular Synthesis' minimalist work produced for the Venice Biennale in 2001. In terms of sound, traces standing for a quality of warmth move past in the background and find their correspondence in the color red. But generally the image world is cool, perhaps even aseptic, the image field is empty, until the granular principle is reaffirmed visibly. The experience is ambivalent and immersive. I think «RESET» is very much where we want to be.

The examples of our work introduced here are sufficient to describe a path I would sum up like this: the theme of the first works was exploring the material, i.e. sound and image, and our fascinationwith various ways of manipulating faces. At first the sound was aggressive and avant-garde, deriving absolutely from a basically attacking position. The theme of the later works is defining space and the image as a light-image, ultimately as beams and modulations within the space; the sound is haptic material, charged and with tension.


Q: How are the samples created, how are the performers recorded?

A: There is no camera movement, the whole thing is a strict installation. The performer acts out emotions ad lib, on the basis of definite requirements, thus creating vowels and consonants. The camera is fixed, the light is white, the background is white, the face or the body are optimally placed in the space. So the image is sterile and stylized, the abstract samples have first been generated and animated with various kinds of image software, and then sound was added in a second step.

Q: Why are the works so big?

A: They aren't always, but it is true that we used to be very interested in massive installations and performances, initially to get away from what we saw as the randomness and object quality of the monitor. We did not want to present our material by saying: this is what we mean, we wanted to stage it as a really impressive situation to be experienced. Seen in this way, the experience was our chief concern. Given that video is a poor pictorial format we came up with the idea of simply moving out into a larger field and using multiple projections to generate almost cinematic image resolution, along with three-dimensional sound. Despite this, we did not intend, or did not intend for long, to wear the audience out. There were protests of course, but that was absolutely ok in the context of being rebellious young artists. Today we sometimes see ourselves confronted with a romantic transfiguration of the early shock shows. We work differently now, and certainly use resources more precisely, focused in aggression.

Q: So is your sound all granular synthesis?

A: No, it varies considerably. On the one hand we have audio-visual material that is entirely coherent within itself. Then, within the framework of anythingthat can work at all on the video grid as granular, that actually is a granular sound-image synthesis. And there are also several acoustic commentary tracks, some created earlier and some performed live. This is handled mainly intuitively, in live situations at least, and as a rule has nothing to do with granular synthesis. As well as this, there is a possibility of disassociating the sound entirely from the image and thus leaving the image alone, which doesn't sound particularly exciting at first, but works very well after an hour's synchronization. You can carry on listening to the sound the whole time, or conversely you can still see the image when the sound is flooding the dark space. We are playing with our core competence, as it were.

Translation by Michael Robinson

© Media Art Net 2004