Path: chuka.playstation.co.uk!news From: Christoph Luerig Newsgroups: scee.yaroze.programming.3d_graphics Subject: Re: polyMorphing? Date: Fri, 11 Dec 1998 11:10:22 +0100 Organization: PlayStation Net Yaroze (SCEE) Lines: 120 Message-ID: <3670EF8E.42AAEFD@immd9.informatik.uni-erlangen.de> References: <366edfaf.411234@news.playstation.co.uk> NNTP-Posting-Host: faui90.informatik.uni-erlangen.de Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="------------A546D217C6FBFCCD6111CF07" X-Mailer: Mozilla 4.04C-SGI [en] (X11; I; IRIX 6.3 IP32) --------------A546D217C6FBFCCD6111CF07 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello! Mesh Morphing seems to be sensible for me only if you are using different levels of resolution for your animation stages. This would be something like the Geomorph Hoppe presented on this years Siggraph. To my guess it is also probably a combination of one and two. You could model this deformation by defining a skeleton, where the different parts of the sekleton are animated using affine transformation by matrices. The skin is then modeled by NURBS patches, where the control points are defined in relation to the skeleton. If a patch is streched over different sekelton parts you get the desired continuity properties depending on the degree of your patch (Poweranimator of Alias|Wavefront has a modeling option that works this way). By tesselation of the patches you get a polygon description that is relative to your skeleton parts. This way you should get fewer necesarry keyframes for your animation than by describing every vertex position in global coordinates. Christoph Yannick Suter wrote: > hi ol! > > I was wondering how the characters in games like crashBandicoot or > tombraider are animated. (?) .. well, I think there are several > techs.. > > (1) single objects, which have to be rotated, transl.. etc > -> character looks quite ugly > (2) every animationStep is saved > -> needs a lot of mem > (3) meshMorphings > -> seems to me the best solution > > are there others? .. which is normaly used? > > a+ Yan > AsC - [code - pixel] - www.netlink.ch/nowadays > Yan - [NetYaroze - cOde] - www.cyratech.ch/dream-y -- ------------------------------------------------------------------- Christoph L"urig cpluerig@immd9.informatik.uni-erlangen.de Graphische Datenverarbeitung Tel.: +49 9131 8529929 Universitaet Erlangen Fax: +49 9131 8529931 cpluerig@immd9.informatik.uni-erlangen.de --------------A546D217C6FBFCCD6111CF07 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hello!

Mesh Morphing seems to be sensible for me only if you are using different levels of resolution for
your animation stages. This would be something like the Geomorph Hoppe presented on this years
Siggraph. To my guess it is also probably a combination of one and two. You could model this deformation by defining a skeleton, where the
different parts of the sekleton are animated using affine transformation by matrices. The skin is then modeled by NURBS patches, where the control points are defined in relation to the skeleton. If a patch is streched over different sekelton parts you get the desired continuity properties depending on the
degree of your patch (Poweranimator of Alias|Wavefront has a modeling option that works this way). By tesselation of the patches you get a polygon description that is relative to your skeleton parts. This way you should get fewer necesarry keyframes for your animation than by describing every vertex
position in global coordinates.

Christoph
 

Yannick Suter wrote:

hi ol!

I was wondering how the characters in games like crashBandicoot or
tombraider are animated. (?) .. well, I think there are several
techs..

(1) single objects, which have to be rotated, transl.. etc
-> character looks quite ugly
(2) every animationStep is saved
-> needs a lot of mem
(3) meshMorphings
-> seems to me the best solution

are there others? .. which is normaly used?

a+ Yan
AsC - [code - pixel] - www.netlink.ch/nowadays
Yan - [NetYaroze - cOde] - www.cyratech.ch/dream-y

  
-- 
-------------------------------------------------------------------
Christoph L"urig          cpluerig@immd9.informatik.uni-erlangen.de
Graphische Datenverarbeitung                  Tel.: +49 9131 8529929
Universitaet Erlangen                         Fax:  +49 9131 8529931

cpluerig@immd9.informatik.uni-erlangen.de
  --------------A546D217C6FBFCCD6111CF07--