On Tue, 26 Jan 1999 23:48:54 GMT, Steve <hor### [at] osuedu> wrote:
>On 13 Jan 1999 09:09:00 -0500, par### [at] my-dejanewscom (Ron Parker) wrote:
>
>>On Tue, 12 Jan 1999 22:20:41 -0600, Mike <Ama### [at] aolcom> wrote:
>>>Looking really good.  When I are we going to get to see how it works?
>>>
>>>I never got around to getting a postscript reader to read Jensen's paper
>>>on photon mapping.  Does it store the lighting values in the screen
>>>space or does it use maps for the objects in the scene?
>>
>>Neither.  It stores them as points in a 3 dimensional lookup structure 
>>independent of any objects in the scene but also independent of viewing
>>angle.  In Jensen's case, he uses a kd-tree.  Actually, Jensen has at
>>least three papers that apply to photon mapping.  All but his SIGGRAPH
>>'98 paper are on his website, and if you email him he'll send you the 
>>URL for that one.
>>
>
>Photon mapping requires that surfaces are divided into finite pacthes.  This
>is, in general, not possible with POV-Ray.   POV-Ray will be most compatible
>with a stored version of monte-carlo.  Monte-carlo "sits" on top of a
>ray-tracer, needing nothing more than to trace more rays.

If you think photon mapping requires finite patches, perhaps you 
should read Mr. Jensen's papers again.  From the bottom of page 8 
in his "Global Illumination Using Photon Maps":

  A very important aspect of the photon map is the fact that it
  is easy to integrate into existing ray tracing programs since 
  it only requires the existence of intersection routines for
  each object.  The scene does not have to be tessellated and
  the photon map structure is completely separated from the 
  geometric representation.

Post a reply to this message