Magmimesis
History
I had the initial insight for this project during the "Specialization in artistic patrimony conservation and the use of new technologies" Media Centre d'Art i Disseny-Mecad's course in 2006. This year I was also in the Mecad's "International Master on Interactive Systems" and one of the subjects was "Theory and practice of interfaces for interactive installations" by the professor Eugenio Tisselli. The aim of this subject was to develop basic skills on physical interfaces and an initial project. "Magmimesis" was my project for this subject.
Theoretical plan
This project has the focus on the conservation problem of digital works of art that in the last decades had used the CRT (cathodic ray tube) type monitors. Nowadays we are living the end of this apparatus period that, following the industrial paradigms, will no longer be produced, being substituted by similar equipments of other technologies like LCD's (liquid crystal display) or plasma. However, the CRT apparatus has the peculiarity to be electromagnetically interfered. These interferences can be seen as an image distortion. This feature was also explored in the art technology dialog of some works like in the example below:
"McLuhan Caged (in Electronic Art II)"
Nam June Paik
In the monitors that starts to replace the CRT's this phenomenon doesn't happen. Still following the industrial paradigms, the apparatus must have a limited lifetime. Thus, we are confronted with the question:What should we do when these devices no longer work? Or how to conserve pieces with planned obsolescence? One possibility is try to handcraft in the artisan way, but this should be an excessively expensive work, in face of the complexities of the production process. Beyond this, the works of art that used the CRT's tried to establish a dialog with technologies and audiences of their epoch. Therefore, once this role had been accomplished, the continuity of working goes from the aesthetic to historic concern, what lead us to imagine another aspect of conservation that should be the preservation of the functionality in opposite of the preservation of the hardware, or, at least, put the question of how much resources should we spend in order to conserve devices that no longer has relationship with nowadays technologies.
The Piece
This piece has two table mounted monitors. One of them is a black and white CRT (cathodic ray tube) and the other one is a color LCD (liquid crystal display). Over the CRT there is a magneto that caus es an image distortion. This magneto is supported by an arm mounted device with potentiometers in the articulations. As the user displaces the arm in order to explore the distortions the changes in potentiometer's angles are electronically read by an arduino (http://www.arduino.cc/) microcontroller and sent to the algorithm that initially converts these readings in a rectangular coordinates po sition. Thus, the algorithm knows the magneto's position over the CRT and uses this information to build mathematically another image with the same distortion. This image is shown on the LCD monitor. The user sees two identical images, one is a product of a physical analogue process, the other one is its digital double.
"Magmimesis aparatus"
Development
First of all I must say that this development only was possible because I was in a scholarship at the KHM (Kunsthochschule für Medien) in Cologne, Germany. I will always be grateful for the helpful and kindness support people gave to me.
Initially I started with a study of the magnetic distortions over the CRT's monitor. I remembered from my childhood to see my older brother distorting the images of our TV set with a 30 centimeters diameter bobbin, but it wasn't the desired effect. I also didn't had any idea about the magnitude of the magnetic field necessary for a good observable distortion. So i needed to try. To do so i built a bobbin using enameled cooper wire in a regular iron screw core. In fact i needed to built a few of these bobbins varying few parameters like the wire diameter and number of loops. The main concern here was to find a good balance between magnetic power and heat dissipation. I tried with these bobbins using a variable direct current power supply. With this equipment you can fix the amount of electric current or the electric voltage, performing small changes with security. After few tests I found a good device for the aims of this project. It produced a good image distortion and wasn't too heat to be handled.
Using processing (www.processing.org) a programed a line grid on the CRT in order to get the first visual approach to the algorithm that I had to develop. This set and the first result is shown in the picture below:
Distortion by an orthogonal magnetic field over a CRT
Great part of my feeling in being able to manage this project was due the years that I spent in the physics institute (IFUSP Instituto de Física da Universidade de São Paulo – Brasil). Of course I couldn't remember how to manage the theories involved in this phenomenon. But with a little research everything becomes clear. The interaction between a moving charge (the electron bean) and a magnetic field (generated by the magneto) is given by the vector cross product (velocity by magnetic field). I will not explain the math here, but with the distortion of the line grid and this equation in my mind I figured out how should be the core of the algorithm.
I started programing for distort just one frame that was a saved image of the same line grid that I used before. After adjust few parameters I achieve a distorted image quite near of the magnetic produced one.
Superposition of magnetic and algorithmic distortions
The next step was to try with a real image. The distortion was produced pixel by pixel i.e. I got every pixel of the font image, calculated the new position and generated a new image. And,of course, the result wasn't what I would like as you can see in the next image:
First distorted image
I realized that I would need to spend a little more time and energy to solve this problem. What was happening was that pixels, that always has integer positions, sometimes got their rounded values jumping one position. Let's say if the algorithm calculated a horizontal position of 9.4 for one pix el it was placed in the ninth position, but if for the first neighbor of this pixel the result was 10.6 then it was placed in the eleventh position leaving one black pixel between them. The solution was to establish a control for every pixel on the screen. If one pixel was filled by the algorithm then it kept its color value, else its color value was given by the average of its neighbors. Obviously this procedure has a cost. In this case this cost was processing time. The algorithm needed to take every pixel, calculate the new position, verify for non-filled pixels and take the average of the neighbors with a video in real time. It didn't work. The distortion was O.K. but the frame rate was less than two frames per second. Completely insufficient.
Talking with the people at KHM I understood that I had to use OpenGL resources. This meant change the software system. My second try was with MaxMSP and Jitter. I had worked with Max before, but never with Jitter. It took me one week of work to translate the software and it worked pretty well. In the same way of Processing, MaxMSP has a very good communication with Arduino board, that was my next focus.
I planed an arm mounted device with two potentiometers in the articulations and the magnet at the end. So, the Arduino just had to take the readings of the potentiometers and send this information to MaxMSP. It worked pretty good.
The arm mount under construction.
Another concern about the construction was with the color CRT. These devices have three electron beans, one for each primary color (red, green and blue). They also have a metal mask just behind the front glass. When a magnet comes near this mask comes magnetized also and produces a distortion in the colors together with geometric distortion. As a general rule this can produce a permanent damage on the CRT. Beyond this, the color distortion depends of numerous factors like details of construction, trajectory of magnet, atmospheric conditions. That is to say, it is a phenomenon with a very complex emulation. To avoid this I decided to use an old black and white CRT with just one electron bean and without the metal mask. On the conceptual plane, this decision was enriching the project. Since the black and white CRT was the first on to reach the public and, thus, could bring a greater historic range into the piece.
Prototype
The first prototype of Magmimesis was built in KHM and had a little presentation just for friends and supporting people on December 21st, 2006.
First version
The first version was made in Barcelona with support of ESDi - Escola Superior de Disseny for the Exploración.electrónica exhibition.
Exhibitions
-Exploración.electrónica – ESDi – Sabadell, 2007
-III JAMM - Jornades d'activitats i mostra multimedia – Centre cultural Golferichs – Barcelona, 2007
-Loop Festival - Barcelona, 2007
Special Thanks
-My wife Carolina Dimitrov.
-My brother Oswaldo Menks.
-Eugenio Tisselli and Claudia Giannetti from Media Centre d'Art i Disseny, Barcelona.
-Paul, Helga, Daniel and Anne Delhougne our dearest hosts in Cologne.
-Martin Nawrath, Urs Fries, Peter Schuster, Martin Rumori, Winfried Haushalter, Bernd Voss and Gerrit Chee-Caruso from Kunsthochschule für Medien, Cologne.
-Vanina Hofman for her friendship and support in the exhibitions of this project.