Ah, my bad... didn't realize it was such a complex thing. Shimms
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Michael Hobson Sent: Monday, 29 November 2004 10:29 PM To: [EMAIL PROTECTED] Subject: [hlcoders] re: false claim of real-time radiosity support in Source Engine At 02:42 AM 11/29/2004, Michael Shimmins wrote: >How on earth can you make such a claim? You think just because it isn't >implemented in HL2 that the Source Engine isn't capable of it? > >Who's to say HL2 exposed 100% of the functionality of the Source Engine. > >Why not wait until we see what can be done with it, before blindly >stipulating that which can not. Dan Partly stated that it is *obviously* a false claim by Valve that they support Real-Time Radiosity in Source Engine on their web site. He can make this statement with an extremely high degree of confidence. It *should be* painfully obvious to any student of computer graphics technology that the claim is false. Why? Because there has to exist a computer executable algorithm to for real-time radiosity before Valve Software or anyone else can possibly have implemented it in any game engine. If such an algorithm would have been published in any one of the several academic journals bearing on this subject, it would be indexed in the ACM (Association for Computing Machinery) Library and available to SIGGRAPH members for examination. There has been no such publication describing any methodology of performing hemi-cube or any other Radiosity computation in Real Time and the present best algorithm requires at least a few tens to a couple hundred iterations of propogating light energy ("radiosity") over a couple hundred thousand hemicubes (for a typical HL 2 level after subdivision of all the polygons into sufficiently small facets) in order to obtain a solution that "looks good". All of this computation would be required for each and every frame rendered and even polygons that are not in the visible set would still have to be included in the radiosity calculation, as they might contribute scattered light to the visible polygons that actually get rendered by means of multiple bounces. That is absolutely beyond power of any existing consumer hardware and could only hope to be be achieved with a very large cluster of machines working on the problem in parallel. Therefore, it may be stated with *very high confidence* that the claim on Valve's web site is false. Besides that, it's not like we've never seen false claims and statements from Valve Software before, either. So all this shock and horror from you fanboys at the report of another one is quite incomprehensible. {OLD}Sneaky_Bastard! email: [EMAIL PROTECTED] Michael A. Hobson icq: #2186709 yahoo: warrior_mike2001 IRC: Gamesurge channel #wavelength _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
