Re: Question About Perfly

New Message Reply Date view Thread view Subject view Author view

Grambo (graham++at++beast)
Wed, 2 Nov 1994 09:58:29 -0800

Hi Marty, I see this anomoly a lot when people are using the alpha component to
"cookie cut textures". On many Apps it looks like the tree has frosting around
the outside. Basically the alpha component extends further in s and t (texture
space) then the color component. A simple fix is in your favorite texture
editors: Adobe photoshop, Coryphaeus, MultiGen, vi (just kidding I hope), etc
editor bring up the intensity (or if color the color components) only and
extend texels out to fully cover where the alpha punches through. This should
eleminate the banding.

The Mak stealth probably bill-boards the trees and does not have them fixed in
position. To do this the alpha would have to be just right.

GB

On Nov 2, 10:02am, Marty Svedlow wrote:
> Subject: Question About Perfly
> The question I have has to do with the appearance of trees when viewed by two
> different viewers (MaK Stealth and Perfly), both of which are based on the
> Performer library.
>
> We have generated a database in Flight format. We inserted trees in the
> database which are fixed in position. The trees are composed of two
> intersecting planes (at 90 degrees from each other) and single planes for two
> levels of detail. All of the tree textures are grayscale textures (i.e.,
> color is not used). As part of the tree texture maps, we included an alpha
> channel that matches the tree boundaries.
>
> The problem we are having is that the trees appear differently when viewed
> using the MaK Stealth viewer (which is based on Performer and uses Perfly)
> compared with using Perfly directly. When using the MaK Stealth viewer, the
> edges of the trees (especially at far ranges) have a heavy black border
around
> their edges. However, when the same database is viewed using Perfly
directly,
> there are no heavy black borders around the edges of the trees. The heavy
> black borders detract from the realism of the scene.
>
> We spoke with MaK Technologies, and they indicated that their may be some
> parameters that need to be adjusted in perfly.cc.
>
> If someone can help to resolve this difference, I would greatly appreciate
it.
>
>
> Thank you very much.
>
> Marty Svedlow
>
>
>
>
>-- End of excerpt from Marty Svedlow

-- 


"Do not follow where the path may lead.  Go instead where there is
no path and leave a trail."


----------------------------------------------------------------------
Graham (Grambo) Beasley		Silicon Graphics, Inc.
MTS (Simulator Guy)		(415) 390-5420		graham++at++sgi.com
----------------------------------------------------------------------

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:50:38 PDT

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