Links of my travel :) - Realtime Raytracing chip, etc

[Richard McCarthy]

A project of making realtime raytracing chip

http://www.saarcor.de/

 

OpenRT Project :

http://www.openrt.de/index.html

 

 

"The MassiveRT Project aims at interactively visualizing extremely large datasets with a software ray tracing approach based on the OpenRT real-time ray tracing engine. This approach makes it possible to interactively render a complete "Boeing 777" model containing more than 350 million individual triangles without any simplifications at several frames per second even on a single PC. "

 

http://graphics.cs.uni-sb.de/MassiveRT/

[Ernest Burden III]

The mind swims at the prospect.

 

Richard, could you summerize what you've learned in your 'travels' for us?

[Richard McCarthy]

hmm ermmm... geh.. the...........

 

(Is this a what's the meaning of life question?)

[Ernest Burden III]

hmm ermmm... geh.. the...........

 

(Is this a what's the meaning of life question?)

 

More a mean of realtime raytracing chips sort ofquestion

[Richard McCarthy]

hmm I don't know, I guess it's more of a prospect of people been able to produce this kind of chip... Perhaps it could be mass produced and be cheap... perhaps... the technology could be extended to realtime radiosity chip (there are some research done on this already, and with hardware implementation it could be done in real-time for sure)

 

I guess I am really interested in their technology of being able to "render millions upon millions polygon without slow down and shortcuts in realtime". I havn't digest all the articles yet, so if you have read it and have some insight please share with us too

[oluv]

realtime raytracing is not that interesting, but realtime gi or radiosity... WOW!

[Richard McCarthy]

You know, there is a branch of radiosity algorithm research called "hierachical radiosity" that seems to be little known or abandoned for some reason.

 

 

http://www.cs.man.ac.uk/~gibsons/gallery_hr.html

 

 

"Finally, this is an image of the main lecture theatre in the Department of Computer Science. The solution was generated in only 192 seconds."

 

192 Seconds! And that was done with the hardware back in 1995.

 

 

I can see that this method can be made realtime by today's standard with advanced branch prediction and culling method, perhaps made even faster by using hardware implementation. (eg. utilising GPU from your graphic card for vector math processing)

 

 

http://www-sop.inria.fr/reves/personnel/Nicolas.Tsingos/

 

This guy have been focus on the acoustics but somehow I found his research more relevent to the real-time radiosity algorithm. The distribution of the sound energy isn't so different from distribution of light energy (maybe in a broader/crude term)

He used to have more linkes and research paper/gallery on his website but somehow I think he tidy it up a bit more so search around, maybe you will find it.

 

This research paper is also interesting :

"Guaranteed Occlusion and Visibility in Cluster Hierarchical Radiosity"

 

http://www.iro.umontreal.ca/labs/infographie/papers/Leblanc-2000-GOVCHR/occlusion.pdf

 

 

 

This research also shows great potentials in combining other method of optimization :

"Visibility Driven Hierarchical Radiosity"

 

http://w3imagis.imag.fr/Membres/Fredo.Durand/PUBLI/Sketch98/results.htm

[Richard McCarthy]

Another one :

"Hierarchical Level of Detail Optimisation for Constant Framerate Rendering of Radiosity Scenes"

 

http://pubs.cs.uct.ac.za/archive/00000048/

 

and

 

http://iiia.udg.es/Simulgen/pub/publications/index.htm

 

It seems from all my searches, most of those links turn up from france..

..odd....