I'm not sure this is helping any or you already know of the caching. I
have working reloadable refence model tails under their own model inside
character. With tons of problems I've faced learning usage of referenced
models I thought I had luck with tails while others have plenty problems
with them. But after reading this I researched a little more and it is not
that simple anymore.
So when I turn cache on and save scene with referenced model tail dies on
next load. If i leave it off tail is alive on load. When tail dies,
instead of live tail mixer has animation SpringCacheSources that I can
load manually on tracks, those seems fit what I left on cache. I armed
all tails with FK controls + skinbones so I can live with this. Spring
simulation is one go after activation of cache + FK. If I've misunderstood
or you have anything to ask please do so, I'm not very confident about the
matter, and will test some more tomorrow.
Mikko
On Tue, 24 Apr 2007 23:15:11 +0300, Steven Caron <carons(at)gmail.com> wrote:
thanks for this.. it is something to go on
steven
On 4/24/07, Ahmidou Lyazidi <ahmidou.xsi(at)gmail.com> wrote:
I have found a way to reconnect the evolve to the cache, if this can
help
someone:
//// select the evolve operator then the cache clip in the sources
////to diconnect the operator:
selection(0). DisconnectGroup( 3, 0, 1);
////to reconnect it:
selection(0).ConnectToGroup( 3, selection(1) ) ;
but in my situation this is not helping so much, beacause the evolve is
not there anymore
Today I tryed 3 differant ways to reconnect things, this is working with
regular models but not with ref:
1- activate and desactivate the mixer animation local modification with
refmod events,this is working in a way, but the animation clips were not
saved anymore
2-deconnect and reconnect the evolve ops with refmod events, but this is
not an operation supported by the deltas
3-having the source model with deconected evolve ops and reconnect them
with a refmod postload event but the operator is removed when I'm
exporting
the model because there is a missing connection.
Regards.
Ahmidou.
2007/4/24, Steven Caron <carons(at)gmail.com>:
>
> i have contacted support and hope we can figure this out. the springs
> work well, until they break. then they just stay broken ( like you
said )
>
> something that others have noticed here at work is the operators seem
to
> update kinematic information but we can't actually see how they are
> connected through the gui. this is upsetting for a td that is trying
to
> figure out how things are working and in this case why they are "not"
> working. i would like to see the "evolve" and "spring" ops properly
> documented, so we can use, distribute, and fix them confidently.
>
> steven
>
> On 4/24/07, Bradley Gabe < withanar(at)stanwinston.com> wrote:
> >
> > I would just assume that once the wiring is broken, the rig is
> > broken. So far there hasn't been a publicized discovery of any
technique for
> > reconnecting the SCOP to its custom Action Source cache node.
> >
> > i am having this same problem in 5.11... i browsed the archives
about
> > this topic and none of the solutions mentioned in those threads
have proven
> > useful.
> >
> > i am thinking about a postload script for the refmodel to rewire
> > anything that has been broken. but i haven't figured out how to
reconnect it
> > interactively. so i dont know where to start programmatically
> >
> > steven
> >
> > On 4/23/07, *Ahmidou Lyazidi* < ahmidou.xsi(at)gmail.com> wrote:
> >
> > Hello !
> > I'm actually testing the new refmodels for our next production and
> > having problems while I'm using them with tails.
> > The fact is that the tails are working the first time, but if I
reload
> > my scene, they are not working anymore.
> > The only way is to disable the animation-mixer modifivation check in
> > the refmodel, which is quite limitative!
> > Did I miss something?
> >
> > PS:The tail was created with the curve inside the model and is
> > working.
> >
> > Best Regards
> > Ahmidou.
> >
> >
> >
>
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
