Re: Non-MultiSample...Q: Transparency
Rémi Arnaud (remi++at++remi.asd.sgi.com)
Wed, 12 Feb 1997 17:59:46 -0800 (PST)
Martin Suter wrote:
>
> According to Angus Dorbie:
> >
> > I was just talking about sorting transparent objects in the scene, against
> > opaque and perhaps against eachother.
> >
> > There's a _vast_ difference between this type of sort and the kind of
> > polygon level sort required by the front to back rendering on the alpha
> > saturate full scene antialiasing you describe. There are also implications
> > for meshing the scene.
>
> Performers,
>
> Let me ask a more general question about transparency here.
> Suppose you have many objects with partially transparent
> textures applied (e.g. a forest made up by many billboards).
> On a RealityEngine2 graphics system there is no need to
> render the transparent geometry back to front to get proper
> results, as long as you use the transparency type PFTR_MS_ALPHA.
>
> This is different for other graphics subsystems. What types of
> SGI graphics subsystems support the benefit of not requiring
> sorting of transparent geometry? High IMPACT, Maximum IMPACT,
> and OCTANE/MXI are of particular interest to me.
>
>
None. You need multisample to have MS_ALPHA screen door transparency.
Note that you need to sort if you use the high quality PFTR_BLEND_ALPHA
transparency.
Performer uses the default bins to do that sorting. (see man pfChannel)
_ / _ _
|_) _ ._ _ o /\ |_)|\ | /\ | || \
| \(/_| | || /--\| \| \|/--\|_||_/
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2
on Mon Aug 10 1998 - 17:54:39 PDT