[Sidefx-houdini-list] Anisotropic work around?

jrutherford at stclaircollege.ca jrutherford at stclaircollege.ca
Sun Feb 19 12:17:29 EST 2006


Hey Peter:
I think that the default anisotropic lighting model is based on Greg
Ward's lighting model which is .... ahhhhh...
anyways for something like brushed metal, you are getting that long thin
specular because the shader is finding the x direction and the y direction
using something like
 vector xdir = normalize (dPdu);
Anyways that works great for a surface but with polygons, the houdini
default gives you that kind of specular on every single polygon which
looks like crap ....but is exactly what the shader is supposed to do...
so what is the workaround and why doesn't Sesi have a default that works
on polys like Maya...hmmm...probably some patent thing...
I think Maya is using some sort of Photon Tracing to get it to work but
I'm probably wrong...
You'll have to talk to the overly caffeinated ex oboe player for that one
which I think in my own humble way is what he has been sort of talking
about in his other thread.
Jim







> Hi,
>
>    I was hoping a kind soul could explain to me in somewhat plain English
> how
> I can render anisotropic highlights when using polygonal surfaces
> containing
> faces with less or more than 4 verts. I have tried to get my head around
> the
> recent topic on sidefx.com, "creating a brushed metal shader". A few
> technical explanations were given in vex code, which I'm embarrassed to
> say
> is completely out of my league... I don't even know how to approach it...
> most of the time I'm able to hack something together... but with vex
> surface
> shaders I'm quite lost. While I would love to take the time to learn how
> to
> program shaders in mantra, it's unfortunately an undertaking that requires
> a
> great amount of time and playing around.
>
> To me personally this is one of the areas in houdini which frustrated me
> the
> most. Take the example of a default anisotropic highlight shader, it
> should
> be able to render out a polygonal surface correctly, without the user
> having
> to dig so deep as into vex code. It baffles me why there is no more
> detailed
> explanation of this in the documentation. The line- "Avoid using the
> Anisotropic Specular model on non-subdivided polygonal geometry because it
> will look flat shaded." This is only partly correct as poly geometry built
> with faces NOT quads, are rendered wrongly by the specular lighting model
> anyway, even with the subd turned on.
>
> Another issue for me is also why the rendered output from a subdivide sop,
> and
> a polygonal model with 'Polygons as Subdivision Surfaces' turned on in the
> render tab has different output. But that's another story...
>
> Anyway.... back to banging my head against the wall... wohoo ;)
>
>
> Peter.
>
>
>
> _______________________________________________
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
> https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
>





More information about the Sidefx-houdini-list mailing list