Announcement

Collapse
No announcement yet.

Unbiased Randering and Caustics

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by DADAL View Post
    Isnt blender render engine open source? And cycles? I thought they were open huhh
    Ahh I see what you mean - yes blender is open but its not the same thing really: I guess I see an engine as something that resolves many things in real-time, not just light transit. I wouldn't describe an offline renderer as an engine for this reason. Brigade is being developed specifically for games development in mind; to resolve a dynamic and changing environment such as that in a game with physics, animation, rigging and so on. Imagine a feature rich real-time engine based around PT rendering such as Brigade on future graphics hardware. That's how awesome the potential is and why I feel an open source approach would be great.
    Last edited by deflix; 22-03-2013, 10:26 AM.
    Immersive media - design and production
    http://www.felixdodd.com/
    https://www.linkedin.com/in/felixdodd/

    Comment


    • #32
      Originally posted by tricky View Post
      When can I have that?!?!
      You can get an old version here - http://igad.nhtv.nl/~bikker/
      For a newer Alpha I suggest contacting OTOY
      Immersive media - design and production
      http://www.felixdodd.com/
      https://www.linkedin.com/in/felixdodd/

      Comment


      • #33
        Here is a good article about OTOY, Brigade and Octane.

        http://www.fxguide.com/featured/octa...e-ray-tracing/

        and a paper - http://www.google.co.uk/url?sa=t&rct...,d.ZWU&cad=rja

        enjoy!
        Last edited by deflix; 22-03-2013, 09:16 AM.
        Immersive media - design and production
        http://www.felixdodd.com/
        https://www.linkedin.com/in/felixdodd/

        Comment


        • #34
          Hi,

          "Photon Parameterisation for Robust Relaxation Constraints" looks pretty cool to remove the noise from caustics.

          EG2013: http://cs.swan.ac.uk/~csbenjamin/zetarelax.html
          (Video): http://vimeo.com/57395372

          Comment


          • #35
            Originally posted by sersia View Post
            "Photon Parameterisation for Robust Relaxation Constraints" looks pretty cool to remove the noise from caustics.
            This is patented last time I looked into it. So forget about it

            Best regards,
            Vlado
            I only act like I know everything, Rogers.

            Comment


            • #36
              Originally posted by vlado View Post
              This is patented last time I looked into it. So forget about it

              Best regards,
              Vlado
              aha! I got it.

              Thank you for your reply.
              Have a nice weekend ahead.

              Comment


              • #37
                Originally posted by vlado View Post
                This is patented last time I looked into it. So forget about it

                Best regards,
                Vlado
                Is the reason you never play with patented tech because of the problems it can create down the line, or is it a principle thing of refusing to pay and get a license for a bit of maths? Does anyone (other engines) play with patented tech?
                Seems to me that there's an awful lot of incredible features which render engines cant use because they're locked up in patents.

                Comment


                • #38
                  Originally posted by cubiclegangster View Post
                  Is the reason you never play with patented tech because of the problems it can create down the line, or is it a principle thing of refusing to pay and get a license for a bit of maths? Does anyone (other engines) play with patented tech?
                  To be fair, the guys that wrote that paper in particular did approach us to see if we were interested in licensing it. I was not convinced it was worth though (I'm still not). I'm not against licensing stuff where we think it could be useful (f.e. we did license MetaSL, although as far as I can tell, that turned out to be money down the drain).

                  Best regards,
                  Vlado
                  Last edited by vlado; 29-03-2013, 10:28 AM.
                  I only act like I know everything, Rogers.

                  Comment


                  • #39
                    Thanks for the info

                    Comment


                    • #40
                      I could see how licensing a technology that would dramatically speed up your rendering over all would be beneficial. But caustics on its own, is such a thing that is not used widely and the small number of people who tend to use them get about getting the caustics done in old fasion ways. This will not make the effort worth wile in my opinion.

                      Im not sure of the purpose of MetaSL is it used in games or is there a general application for rendering with it? what would be benefits of using it?
                      Last edited by Morbid Angel; 29-03-2013, 12:47 PM.
                      Dmitry Vinnik
                      Silhouette Images Inc.
                      ShowReel:
                      https://www.youtube.com/watch?v=qxSJlvSwAhA
                      https://www.linkedin.com/in/dmitry-v...-identity-name

                      Comment


                      • #41
                        Originally posted by Morbid Angel View Post
                        Im not sure of the purpose of MetaSL is it used in games or is there a general application for rendering with it? what would be benefits of using it?
                        It was mostly because we were often be asked if one could write shaders for V-Ray easily and MetaSL was a way to get a shading language. But it did not get developed...

                        Best regards
                        Vlado
                        I only act like I know everything, Rogers.

                        Comment


                        • #42
                          Here is a small test of the current vertex merging implementation.

                          Click image for larger version

Name:	vertex_merging.jpg
Views:	1
Size:	404.3 KB
ID:	847241

                          Best regards,
                          Vlado
                          I only act like I know everything, Rogers.

                          Comment


                          • #43
                            And another one:

                            Click image for larger version

Name:	bpt_cornell1.jpg
Views:	1
Size:	427.7 KB
ID:	847247

                            Best regards,
                            Vlado
                            I only act like I know everything, Rogers.

                            Comment


                            • #44
                              A comparison between regular bidirectional path tracing and bidirectional path tracing with vertex merging for the same amount of render time.

                              Pure bidirectional path tracing:
                              Click image for larger version

Name:	bpt_no_vm.jpg
Views:	1
Size:	62.9 KB
ID:	847248

                              Bidirectional path tracing with vertex merging:
                              Click image for larger version

Name:	bpt_vm.jpg
Views:	1
Size:	57.1 KB
ID:	847249

                              Best regards,
                              Vlado
                              I only act like I know everything, Rogers.

                              Comment


                              • #45
                                And one more...

                                Click image for larger version

Name:	bpt_cornell3.jpg
Views:	1
Size:	431.8 KB
ID:	847250

                                Best regards,
                                Vlado
                                I only act like I know everything, Rogers.

                                Comment

                                Working...
                                X