TransAtLab - New Atlantis
“New Atlantis” is the code name for a 3-year research project by the TransAtLab program (see below). It is a multi-user cross-platform open source online world inspired by Second Life, adding audio work possibilities. I have participated at two of the project's workshops : in the Aix-en-Provence School of art, from 5 January until 16 January 2009, and in the School of the Art Institute of Chicago, from 15 of October until 31 of October 2009.
During this two editions we mainly built up an early version of the client-server. Besides the core development activities, there were some discussions on the world's properties (interactions, avatar representation, size, 2D or 3D space, etc.) but also about the spacial and sound structure. Some sample object were also created (3D models, Pure Data patches). The client-server are built using Python, Panda3D and Pure Data (the release version will use Pure Data in nGUI mode).
My participation in this research involves the path model for the sound (how it goes from the source to the listener via different spaces and apertures). The pathfinding algorithm is not yet implemented but it is under development for being released in early 2010. The algorithm will use invisible objects included in the 3D model file in order to get knowledge of the geography of the environment. This invisible objects define walls and apertures. For CPU power saving purposes, some of the distances are pre-calculated when the model is first loaded in the world. This distances are constants and will not change once the model is definitively installed in the world. There are two levels of granularity for the pathfinding. One where the building is represented by a cross and a second one where the building is fully represented. This is also made for CPU savings. The sound cannot pass two times in the same aperture or room. The sound has a maximum default distance before fading out and al possible paths are calculated. The calculated paths trigger then the different Pure Data patches with attenuation factors (proportional to the distance). This are the basic rules for the first version of the pathfinding algorithm. A future version is possible where the sound also suffers fade out by being forced to turn around corners.
My learning within this project are strongly related to pathfinding algorithms, but also to python programming, basic understanding of Panda3D and 3D modeling programs (such as 3D Studio Max, Maya, Blender, among others), basic understandings on game engine development, sound physics, creation of a virtual model in order to simulate a physical phenomena. This project also allowed me to bring reflexions on sound and virtual worlds as a materials for creative purposes. This project also involves a series of questions regarding the creation of tools for creativity purposes.
Workshop participants :
***October 2009 : Peter Sinclair, Ricardo Garcia, Gonzague Defos de Rau, Peter Gena, Robb Drinkwater, Ben Chang, Margarita Benitz, Charlotte Benedettini, Mathias Isouard, students from the School of the Art Institute in Chicago and myself.
***January 2009 : Peter Sinclair, Ricardo Garcia, Gonzague Defos de Rau, Jerome Joy, Roland Cahen, Brett Balogh, Alejo Duqué, Scott Fitzgerald, Julien Clausse, Anne Roquiny, Peter Gena, Robb Drinkwater, Ben Chang, Margarita Benitz, Brett Balohg, Charlotte Benedettini, Bastien Vacheron, Michael Barret, Matt Griffin, Ping-Yao Chen, Joseph Grimm and myself.
More about TransAtLab :
Extracted from TransAtLab's website.
{image 3}
During this two editions we mainly built up an early version of the client-server. Besides the core development activities, there were some discussions on the world's properties (interactions, avatar representation, size, 2D or 3D space, etc.) but also about the spacial and sound structure. Some sample object were also created (3D models, Pure Data patches). The client-server are built using Python, Panda3D and Pure Data (the release version will use Pure Data in nGUI mode).
My participation in this research involves the path model for the sound (how it goes from the source to the listener via different spaces and apertures). The pathfinding algorithm is not yet implemented but it is under development for being released in early 2010. The algorithm will use invisible objects included in the 3D model file in order to get knowledge of the geography of the environment. This invisible objects define walls and apertures. For CPU power saving purposes, some of the distances are pre-calculated when the model is first loaded in the world. This distances are constants and will not change once the model is definitively installed in the world. There are two levels of granularity for the pathfinding. One where the building is represented by a cross and a second one where the building is fully represented. This is also made for CPU savings. The sound cannot pass two times in the same aperture or room. The sound has a maximum default distance before fading out and al possible paths are calculated. The calculated paths trigger then the different Pure Data patches with attenuation factors (proportional to the distance). This are the basic rules for the first version of the pathfinding algorithm. A future version is possible where the sound also suffers fade out by being forced to turn around corners.
My learning within this project are strongly related to pathfinding algorithms, but also to python programming, basic understanding of Panda3D and 3D modeling programs (such as 3D Studio Max, Maya, Blender, among others), basic understandings on game engine development, sound physics, creation of a virtual model in order to simulate a physical phenomena. This project also allowed me to bring reflexions on sound and virtual worlds as a materials for creative purposes. This project also involves a series of questions regarding the creation of tools for creativity purposes.
Workshop participants :
***October 2009 : Peter Sinclair, Ricardo Garcia, Gonzague Defos de Rau, Peter Gena, Robb Drinkwater, Ben Chang, Margarita Benitz, Charlotte Benedettini, Mathias Isouard, students from the School of the Art Institute in Chicago and myself.
***January 2009 : Peter Sinclair, Ricardo Garcia, Gonzague Defos de Rau, Jerome Joy, Roland Cahen, Brett Balogh, Alejo Duqué, Scott Fitzgerald, Julien Clausse, Anne Roquiny, Peter Gena, Robb Drinkwater, Ben Chang, Margarita Benitz, Brett Balohg, Charlotte Benedettini, Bastien Vacheron, Michael Barret, Matt Griffin, Ping-Yao Chen, Joseph Grimm and myself.
More about TransAtLab :
TransAtLab is a program of collaborative research and educational exchange in electronic art, including sound, interactivity, immersive environments, telecommunications, kinetics and robotics. Founded in 2005 as a partnership between the School of the Art Institute of Chicago, the Ecole Superieure d’Art d’Aix-en-Provence and the Ecole Nationale Superieure d’Art Villa Arson, Nice, the project brings together the research efforts and expertise of the three institutions in a common academic network. Components of the project include student exchanges, faculty exchanges for lectures and workshops, long-term collaborative research projects in electronic art, and research in the use of telecommunications and hypertext technologies to support distance learning and the development of shared curriculum initiatives.
Extracted from TransAtLab's website.