What’s interesting, though, is that now it has a few thousand views, YouTube is willing and able to provide some ‘insight’ statistics on it:

I find the idea of kinetic sculpture being particularly popular in Ethiopia distinctly gratifying.

I also note the close inverse correlation between me speaking and Hot Spots (or in this case, ‘less cold spots’, perhaps).  Not at all unreasonable in the circumstances.

Abstract as posted on the Goldsmiths Computing Mutators Research Group site:

EMBO Workshop on Visualizing Biological Data

3-5 March 2010, EMBL Heidelberg, Germany

FoldSynth is an interactive multimedia platform designed to help people understand the characteristics of protein structures. It uses a simplified model of molecular forces to give an interactive animation of the folding process. The shape of the protein is shown as a 3D visualization floating above a 2D triangular distance matrix [3] which provides an alternative interactive visualisation. The user adds and removes forces by interacting directly with the 3D protein or by “painting” on the 2D matrix. The shape of the protein reacts to the current set of forces with corresponding change to the distance matrix. There are simple tools to paint forces that e.g. cause the creation of alpha helices and beta sheets. Cellular automata can also operate on the distance matrix, introducing dynamic force processes. The forces can be initialized to mimic proteins loaded from a DAS server. The model includes masses and simplified forces for hydrophobicity and electric charge. The 3D visualization shows various representations (smoothed backbone, cartoon, etc.). Forces and properties (such as hydrophobicity) are mapped onto different graphical features (colour, texture, bump mapping) and sonifications. FoldSynth can recreate real protein structures, but can also create new folds which are protein-like yet fundamentally different to know protein structures. FoldSynth is a continuation of previous work we did on reading DNA data to produce visualisation of protein information which we first presented at SIGGRAPH in 2007. In this previous work we introduced a novel way to navigate backward and forward in time through the phylogenetic tree [1,2]. References: [1] “Using DNA to Generate 3D Organic Art Forms,” W. Latham et al., EvoMusArt 2008. [2] “From DNA to 3D Organic Art Forms,” W. Latham et al., SIGGRAPH 2007. [3] “Protein Geometry, Classification, Topology and Symmetry”, Taylor & Aszodi, IoP, 2005.

Inspired by Nick Didkovsky’s “Density Trajectory Study II For solo guitar and computer“ I’ve recently been spurned into getting hold of a few alligator clips for the purposes of preparing guitar strings.  Placed on modal points of the strings, they exchange the simple harmonics of an unprepared string for beautiful and complex gamelan-like timbres.

Here is a recording of some (sketchy) improvisation, playing a thus-prepared guitar laid on a desk with a pair of pencils.

Playing around with a big glass jar in my bath.  I was going to edit this, maybe in a rhythmic way… then I thought maybe it would be fun to play around with the YouTube Javascript API to do some interactive web browser based editing.  Well, I may never get ’round to that, but here is some noise anyway.

These are 3D structures that can naturally evolve in real-time, or, with the simulation frozen, be viewed from arbitrary angles as static virtual sculptures.

In these images, the ‘seed’ that is fed into the feedback is simply the plain black or white background of the window.  There is a single rectangle rendered with the output of the previous frame as a texture, which is fed through a GLSL shader which inverts colour and applies a kaleidoscope effect, resulting in rich chaotic imagery.

I hope these are enjoyed by anyone who happens to stumble across them.  At some point, I should update with more images, some usable software, video and better explanation.

Click here to download and run the java webstart application.

Instructions

Move the mouse over different parts of the grid to register interest in particular configurations.
The borders fade from green to red, depending on the amount of attention they are given. When they go far enough into the red, they are replaced with new offspring based on the others that are currently around. Influencing the evolution of the population in this way is the main mode of interaction.

Additionally, moving the mouse over the large view to the left causes the evolution process to pause, such that the form may be examined without the pressure of the current population changing. Holding the control key while moving the mouse holds the currently focussed specimen, so that you don’t have to carefully dodge with the mouse if you don’t want the large view to change.

Each individual view can also be manipulated by dragging the mouse:

• Left click & drag for rotation.
• Right click & drag for translation.
• Middle click & drag to scale.
• Hold left & right buttons and drag to rotate about z.
• Alt & click to reset view.

There are no alternatives for mice with less buttons for now; sorry. All comments are appreciated.

Photographing images such as the ones above is a unique experience whereby one can compose a close up view of a particular feature that would normally be fleeting.  So while the photographs may appear very similar to any photograph of splashing water taken with a flash, they are really something very different.

I’ve recreated a similar setup again and made this video, although it doesn’t document all of the interesting features of the real phenomena; I’d love to encourage people to recreate it for themselves, ideally.

Another similar setup, which may well have been Lee’s original inspiration, is described (and also well photographed, along with many other interesting phonema) in Hans Jenny’s “Cymatics” study of wave phenomena, Volume II chapter 6:

The outflow pipe is excited by sound and this causes the emergent water jet to oscillate.  As it transmits periodic impulses it receives into space, the interplay of a number of factors can be seen.  Surface tension causes the jet to maintain a certain homogeneity.  Where it is disrupted, the surface tension is evident in droplet formation.  The main feature, however, is the enormous turbulence apparent in the periodic movement due to the high instability of the jet.  This periodicity also appears wherever the jet is disrupted.  The instability of this interplay of forces can be seen very clearly with the stroboscope.  For a brief moment individual drops have an incipient regularity of formation, but in an instant become turbulent again.

Curious processes involving attenuations and expansions can be seen.  A spiral current appears in the jet time and again.  In spite of the complex instabilities the pattern of oscillation still remains prominent.  It involves a number of tendencies combined together: there is a tendency for the water to turn into drops, to form a homogeneous jet, and to flow in wave patterns, and at the same time a tendency to produce a regular vibrational pattern.  None of these tendencies is ever fully realised; as soon as ay single one becomes more precisely defined, it is caught up in the turbulence caused by the prevailing instability.

In the variations Lee and I set up, a sine wave was used rather than periodic impulses, and I suppose this may have affected a somewhat greater stability than that described by Jenny.  Certainly, there was a sufficient stability that to an observer individual droplets appeared either suspended in air for some time; or slowly ascending or descending in cases where the frequency at which the strobe caught them was higher or lower than the fundamental of the excitation.

The synthesis is “cascaded waveguide”, excited by an audio recording and controlled by midi keyboard.  As such it has some acoustic-ish dynamics, twisted in a way that wouldn’t be physically possible and imbued with some intuitive chance elements, particularly in how the input sound articulates whatever is being played on the keyboard.