secondary ray 2005-03-31 - By Halfdan Ingvarsson
Back
Well.. I believe it's due to the fact that mi_trace_transparency can now be
used with scanline rendering and rapid motion (in a manner of speaking) without
triggering a ray cast and associated BSP tree generation/lookup (if the eye ray
isn't modifed by a lens shader). So it changed from a 'ray' to a part of a
compositing operation, but only if preceded by an unmodified eye ray. Go figure
.
Personally, I also think it was a little bit odd.
- �
> -- --Original Message-- --
> From: owner-xsi@(protected)
> [mailto:owner-xsi@(protected)]On Behalf
> Of Alan Jones
> Sent: Thursday, 31 March, 2005 10:54
> To: XSI@(protected)
> Subject: Re: secondary ray
>
>
> Strange that they changed that - If you wanted it a ray to remain what
> it was wouldn't you just use mi_trace_continue and not bother having
> another state created?
>
> Cheers,
>
> Alan.
>
>
> On Thu, 31 Mar 2005 10:35:59 -0500, Halfdan Ingvarsson
> <hingvars@(protected)> wrote:
> >
> > Exactly. mental images changed the semantics of the
> transparency ray around
> > version 3.2 (from a secondary to a primary ray) but the
> shader did not get
> > updated to reflect this.
> >
> > - �
> >
> > -- --Original Message-- --
> > From: owner-xsi@(protected)
> [mailto:owner-xsi@(protected)]On Behalf Of
> > Jim Rothrock
> > Sent: Wednesday, 30 March, 2005 21:05
> > To: XSI@(protected)
> > Subject: Re: secondary ray
> >
> > Gerbrand Nel wrote:
> >
> > still feels like it shouldn't work this way, if I make
> something 100 %
> > transparent but still have highlights running on it,
> everything behind that
> > object should be a secondary ray... because it's a refraction on the
> > transparent object, even if the index of refraction is 1
> ??? does anyone
> > else agree????? the funny thing is it starts working the
> moment you change
> > the index of refraction. usefull sometimes but isn't that
> what the sprite
> > shader is for????Without having the source code to the
> shader you are using,
> > I can make a guess about what is happening. If the
> relative index of
> > refraction is not 1.0, the ray direction must change, so
> the shader calls
> > mi_trace_refraction(). If the relative IOR is 1.0, the ray
> direction will
> > not change, so mi_trace_transparent() is called instead.
> This is a speed
> > optimization. mi_trace_transparent() will have
> approximately the same
> > effect as mi_trace_refraction(result, state, &state->dir),
> but transparency
> > rays are considered to be primary rays. From the mental
> ray documentation:
> >
> >
> >
> > miRAY_PRIMARY(raytype)
> > is true if the raytype is a primary or a transparency ray.
> > -- Jim Rothrock | Stan Winston Digital | jimr@(protected) ---
> > Unsubscribe? Mail Majordomo@(protected) with the
> following text in body:
> > unsubscribe xsi
>
> ---
> Unsubscribe? Mail Majordomo@(protected) with the following
> text in body:
> unsubscribe xsi
>
---
Unsubscribe? Mail Majordomo@(protected) with the following text in body:
unsubscribe xsi
|
|
