Re: Transparency

New Message Reply Date view Thread view Subject view Author view

John Rohlf (jrohlf++at++tubes)
Mon, 2 Oct 95 14:21:28 PDT


>
> I have seen similar problems when loading Multigen models which
> contain transparent parts (though i'm sure it will happen with
> other models, too). I think the problem is the order of rendering.
> I don't think Performer does any pre-sorting of transparent parts,
> and to get transparent parts to render correctly, they should
> probably be rendered after all opaque parts and sorted to render
> back to front. The MultiGen modeller allows priority to be applied
> to parts, this can change the order of rendering. I'm sure that
> something within Performer will let you assign higher priority
> to your transparent parts so they can be drawn last. I just don't
> know if you can sort them easily.
>
> Thom
> <---- Begin Included Message ---->
> Date: Fri, 15 Sep 1995 16:13:36 -0400
> From: Lynda D Myers <ldmyers++at++afit.af.mil>
> Subject: Transparency
> To: info-performer++at++sgi.com
>
> I am having touble seeing one transparent
> object through another when rendering objects
> in performer. Using the no_occlude option in
> pfTransparency did not solve the problem.
> Is there another way to do this?
>
> Thank you,
> Lynda D. Myers
> e-mail
> ldmyers++at++afit.af.mil
>

        In 1.2, there was a bad interaction between pfLayers/pfDecal
and NO_OCCLUDE so you may be experiencing this with transparent layers.
Another situation is specific to multisampled, screen door transparency.
2 objects of equal transparency will use the same screen door so
even with NO_OCCLUDE you will only see the last-drawn transparent object.
To test this you can explicitly request PFTR_BLEND_ALPHA transparency
or vary object transparencies.

        Performer 1.2 does draw transparent objects after opaque and
2.0 goes a step further by sorting transparent geosets back to front.

         


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:51:56 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.