Looking back ... After working a while with POVRay, I started occupying myself with Radiance in the final months of the last millenium. Since these days Greg Ward (Anyhere Software), the main developer of Radiance at LBNL (California/USA) and EPFL (CH), has in fact become a teacher for me. Besides this or that glance into textbooks I learned C programming by delving into the Radiance code, thereby coming across so many subtleties and valuable insights hidden in the lines. It was and still is simply amazing, and I have to thank Greg a lot for sharing all this wisdom !! It quickly became clear that image generation, especially still images produced with artistic background was the main focus of my interest, and in turn I started to implement this or that addition within classic Radiance to make it better suit my demands. During an engaged discussion about Radiance and Radiance development at the end of 2003 Erwin Rol spontaneously threw in the idea of an independent development branch including a name for the game: 'Radzilla'. Together with my longing to step beyond standard C and learn more about programming, other languages, tools, concepts, etc, this equally spontaneously gave birth to the project. Right from the start it was and is an aim of Radzilla not only to collect all my own Radiance related work, but also to integrate other third party modules, and to link it with external libraries and toolkits (e.g QT for the GUI). There are many contributors out there (some may even not know that they in fact have become one), who helped in one or the other way, e.g. by providing direct feedback, doing some testing, granting access to equipment or just by publishing some valuable info on the net or on Peter Apian-Bennewitz' Radiance website www.radiance-online.org and its mailing lists. I simply want to say thank you all for that!!

Excursion: Visualization vs. Physical Accuracy The fascinating aspect of Radiance is, that it can be used (and this is actually done so, worldwide) for a wide range of different tasks ranging from pure scientific analysis of all sorts of lighting/vision related topics to free 3D art projects. What better compliment can one make about a tool as that it proved itself to be valuable even for applications the original developers haven't thought of at the beginning? Somehow in the middle of this range the topics architecture and design visualization can be located, which both demand a certain amount of accuracy and realism and an aesthetical approach at the same time. In short words, classic Radiance takes care of the accuracy part of the matter, whereas the Radzilla extensions/modifications introduce ways to operate more freely with the functionality provided by the original program. The topic of physical/scientific accuracy always appears in conjunction with Radiance, and undoubtedly Radzilla introduces more ways to 'violate' this aspect, dependent, however, on the users free will to exploit them. No matter how one looks at it, a visualization cannot be judged by the criterium of accuracy alone. Indeed, physical accuracy and validation is a very difficult topic, which quite generally can be strictly adressed only in limited environments (in the extreme: a box and a point source. John Mardaljevic or Roland Schregle can tell you more about this) In contrast, complex architecture visualizations often work on the basis of the much more generalistic terms 'realism' or 'realistic appearance', if they not deliberately distance themselves even more form these thoughts, e.g in order to emphasize central design charateristics. And, just for comparison, there has been architecture and visualization in ancient times, too. How physically accurate were the ink or pencil drawings or maybe even oil paintings former architects and designers presented to the kings and queens as proposals for new castles, churches, museums? Still marvellous buildings with excellent interior lighting have been built. Todays main visualization tool is the computer (although the others still exist and are used), but the principal ideas remain the same. To give just one example: There's always the topic of abstraction in conjunction with images, an image is always something different than the real 3D scene. But if the viewer can grasp the abstraction and make up the connection with the reality to which the abstraction refers, one of the main tasks of a visualization is fulfilled. Carsten Bauer