Some time ago, I asked about the mgf2inv tool, as this is the only program I know that is able to export from radiance's scene format to other 3d formats. However, as the only way goes through mgf, which AFAIK didn't change in the last years, it is e.g. impossible to export the image map information from radiance. I wonder if there is really no need for that?
I found an interesting project called xrml which is enhancing vrml97 into so-called PhBvrml, a vrml extension for physically based rendering. There are even converters for mgf, so again I can get into this format through mgf only.
I think that a similar approach like PhBvrml, simply not based on vrml97 but the xml'ed x3d, would be a great solution to keep scenes in a more readable format also useable by other tools (e.g. x3d-browsers, xml-editors and validation tools). I think that if radiance could use something like that, we would make a big step in the integration of the simulation with other developments.
So, finally, my question is: is there any interest and chance that we will see radiance (directly or with the help of filters, like the mgf-support) will be able to handle such a format that can be read by other browsers and has the extendibility and validation features of xml? Might such a format even be an approach to get into real-time and hw-accelerated radiance rendering? If I can output "prerendered" triangles with e.g. the ambient calculation done in radiance, I could use whatever (accelerated) x3d-browser, couldn't I? So the task to display would be given to other, specialised programs, while radiance would export into standard formats.
I hope my ideas are not to stupid in the context of radiance's use... and on this list. I am just fighting with problems of documentation / model integration, converters and loss of information here...
Lars O. Grobe, firstname.lastname@example.org, ++90-212-2458330
Kardesler Apartment, Turnacibasi Sokak 28
Galatasaray - Beyoglu, 34433 Istanbul