Re: Re[2]: constraints vs viewport refresh/update

["Ahmidou Lyazidi" <ahmidou.xsi(at)gmail.com>]

In my test, it was 20/30% than a constraint

2008/3/10, Frank Lenhard <franky(at)ixdream.com>:

doesnt using op's slows down the update times considerably?
would not hiding the constrained nulls but setting their primary_icon
to 0 instead, help?

ciao
franky


Saturday, March 8, 2008, 7:36:11 PM, you wrote:

j> I havent used constraints in a while because I use operators. I have
j> inquired about this happening in the past and this is done for 'efficiency'
j> which makes sense IF your entire tree is hidden.
j> I looked at your scene and you are right the constraint should be pulled by
j> the necessity of deforming the mesh, somehow the visibility. So this is a
j> bug, and it ought to be logged, yet constraints were written so long ago and
j> softimage at this point has its hands full so i doubt that this will be
j> fixed any time soon.

j> I tried replacing your constraint with an op

j> Delete your constraint, replace it with
j> AddScriptedOp( "deformer1.kine.global","function
j> TEMP_Update(ctx,Out,InG0,InG1){Out.value.transform = InG1.value.transform;}",
j> "deformer1.kine.global,Ctrl_R.kine.global", "TEMP", "JScript", 0)

j> And it works. What this op doesnt do is the compensation, whicsh you would
j> have to work out yourself, but it isnt so hard. all you have to do is cache
j> the initial transform difference (Map deformer global transform to Local in
j> the space of the constrainer), cache that somehow , either extract the
j> values pos, rot, scl, stick tehm into parameters and then inthe update
j> reconstruct that transform from parameters and rebuild that transform., then
j> map that transform to global in the space of the constrainer transform
j> (InG1) and you got it.That is one way.

j> hope this helps






j> On Fri, Mar 7, 2008 at 7:36 PM, Robert Moodie <rmoodie(at)buzzimage.com> wrote:

>>  Yep, some of the character rigging I did in 300 was intense - to final
>> anim to send to lighting I actually wrote a script for the Immortals which
>> stepped through the timeline and saved a key on all the deformer objects.
>>
>> It was literally a manual 'Plot constrained transforms' and there were a
>> couple of places where I had to force a 'Refresh();'. It was a 10 -15 min
>> operation but concrete.
>>
>> The worst cases I have seen are where you have a null(s) constrained to a
>> bone(s), these nulls are cluster centre deformers for curve points. On the
>> curve you have a null which is path constrained and is the actual envelope
>> deformer -- "guided deforms" TM ;-).
>>
>> Great system, but the dependency/refresh wasn't funny.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> -----Original Message-----

>> From: Adam Sale [mailto:adamsale(at)shaw.ca <adamsale(at)shaw.ca>]
>> Sent: Fri 07/03/2008 9:30 PM
>> To: XSI(at)Softimage.COM
>> Subject: Re: constraints vs viewport refresh/update
>>
>> I have been getting this same behavior for a few versions now where after
>> an Apply Action, constraints don't refresh, the only workaround I found was
>> to go back and forth a frame in the timeline..
>>
>>
>>
>> ----- Original Message -----
>> From: Thiago Costa <thiagocosta3d(at)gmail.com>
>> Date: Friday, March 7, 2008 5:29 pm
>> Subject: constraints vs viewport refresh/update
>> To: XSI(at)Softimage.COM
>>
>> > Isn't weird that XSI doesn't update correctly constraints and
>> > other operator
>> > values when the object is not visible in the viewport?
>> > Is there a way to force XSI to read that specific object even
>> > when it's not
>> > invisible?...
>> >
>> > If anyone wanna try... open this scene attached, and play.
>> > You'll see the
>> > animation is all right.
>> > Now hide all nulls or just disable the visibility for nulls in
>> > the viewport
>> > and play it again.
>> > Here it just don't update the nulls animation correctly when you hide
>> > nulls... Is there a performance reason for this be like that?
>> >
>> > It should update operators when they are necessary for Envelops
>> > or other
>> > things that are being displayed... shouldn't it?
>> > Now Raff will complain that I'm attaching scenes too :P
>> >
>> > anyway... it's almost weekend in the pacific... :P
>> >
>>
>>


---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi