xsi 2013 merging envelope operators
-
- Posts: 52
- Joined: 06 Jul 2012, 18:08
- Location: bremen, germany
xsi 2013 merging envelope operators
hi,
i transfered some envelopes with gator from clusters.
now i have multiple envelope operators and envelope clusters.
how can i merge the multiples envelope operators and weights to only one for further modifications?
i transfered some envelopes with gator from clusters.
now i have multiple envelope operators and envelope clusters.
how can i merge the multiples envelope operators and weights to only one for further modifications?
The Soulcage Department
http://www.soulcage-department.de
http://www.soulcage-department.de
-
- Moderator
- Posts: 754
- Joined: 25 Nov 2009, 01:41
- Contact:
Re: xsi 2013 merging envelope operators
I don't know that you can honestly. Would love to be told different. I believe this is a newer feature added in 2013 so not 100% sure.
-
- Posts: 52
- Joined: 06 Jul 2012, 18:08
- Location: bremen, germany
Re: xsi 2013 merging envelope operators
maybe with ice? but how?
The Soulcage Department
http://www.soulcage-department.de
http://www.soulcage-department.de
Re: xsi 2013 merging envelope operators
Gator implies your second mesh haven't any envelope. Theoretically, you can write a script which blends the envelope values between the two and put them into one operator. Maybe even with use of ICE.
-
- Moderator
- Posts: 754
- Joined: 25 Nov 2009, 01:41
- Contact:
Re: xsi 2013 merging envelope operators
I'm not sure I understand what you're saying here. You could always GATOR even if it has an envelope on the source or target mesh. You can now transfer multiple envelopes to the target mesh from multiple objects with multiple GATOR ops.iamVFX wrote:Gator implies your second mesh haven't any envelope.
http://softimage.wiki.softimage.com/ind ... 9#Modeling
Merging them is the issue and could be done through scripting. I don't know that you can do that through ICE (only) as you'd have to add all of the deformers to the target mesh and then get the 2 different envelope maps which have different indices for the deformers. It would probably be easier through scripting though you may need to compensate for any overlapping points within each cluster.
-
- Posts: 52
- Joined: 06 Jul 2012, 18:08
- Location: bremen, germany
Re: xsi 2013 merging envelope operators
i tried to do it with ice:
i duplicated the mesh with the multiple envelope operators, deleted any envelopes on the duplicate and assigned a new envelope with the deformers.
then i built an ice tree wit a "closest location" and added the weights from the clusters together and set it to the weights cluster of the duplicate.
but this gives me very unespected result. the mesh doesnt deform as the original.
maybe someone give me a hint how to blend the multiple envelopes together?
i duplicated the mesh with the multiple envelope operators, deleted any envelopes on the duplicate and assigned a new envelope with the deformers.
then i built an ice tree wit a "closest location" and added the weights from the clusters together and set it to the weights cluster of the duplicate.
but this gives me very unespected result. the mesh doesnt deform as the original.
maybe someone give me a hint how to blend the multiple envelopes together?
The Soulcage Department
http://www.soulcage-department.de
http://www.soulcage-department.de
-
- Posts: 52
- Joined: 06 Jul 2012, 18:08
- Location: bremen, germany
Re: xsi 2013 merging envelope operators
ok.I'm not sure I understand what you're saying here
i have a character that contains multiples meshes(tshirt, arms, head, trousers...).
now i have modelled a jacket (hires). i used gator to transfer envelopes from the differet character parts to the jacket by selecting points (like jacket-arms, jacket-spine, etc).
after that i ended with multiple envelope operators and weight clusters. but now i want to paint some smoothness, weights or apply smooth operators on the whole jacket. so i need a merged envelope operator.
also for animation, multiple operators will slow down the realtime performance...
The Soulcage Department
http://www.soulcage-department.de
http://www.soulcage-department.de
-
- Moderator
- Posts: 754
- Joined: 25 Nov 2009, 01:41
- Contact:
Re: xsi 2013 merging envelope operators
@nosplitters: my quote and first question in my last reply was for iamvfx.
Re: xsi 2013 merging envelope operators
Can't believe they're serious about multiple envelope operatorsEricTRocks wrote:http://softimage.wiki.softimage.com/ind ... 9#Modeling
How they think people supposed to work with weights...Each GATOR operator adds a separate envelope operator.
-
- Posts: 52
- Joined: 06 Jul 2012, 18:08
- Location: bremen, germany
Re: xsi 2013 merging envelope operators
i did it another way:
merged all my character parts to one mesh and transfered the envelope. then i used that merged mesh as the input for gator to weight my jacket.
works. one operator...
thanks for your support.
merged all my character parts to one mesh and transfered the envelope. then i used that merged mesh as the input for gator to weight my jacket.
works. one operator...
thanks for your support.
The Soulcage Department
http://www.soulcage-department.de
http://www.soulcage-department.de
-
- Moderator
- Posts: 754
- Joined: 25 Nov 2009, 01:41
- Contact:
Re: xsi 2013 merging envelope operators
I've done it many times before for various things and its better to be able to do that than not. I'd rather have the option than not.iamVFX wrote:How they think people supposed to work with weights...
Re: xsi 2013 merging envelope operators
True.EricTRocks wrote:I've done it many times before for various things and its better to be able to do that than not. I'd rather have the option than not.iamVFX wrote:How they think people supposed to work with weights...
Re: xsi 2013 merging envelope operators
Can you please explain that in some more detail, I'm not following your steps. Very interested though, I'm having exactly the same issue with character, clothes, etc in separate meshes with separate envelopes. And I couldn't figure out a way to fix...
thanks, Jasper
thanks, Jasper
- Kerro Perro
- Posts: 306
- Joined: 20 Mar 2011, 01:01
- Location: Leiden, the Netherlands
- Contact:
Re: xsi 2013 merging envelope operators
Hey since the original problem is resolved can i ask a question on this subject?
Can anyone (cough..like Eric..cough ;) ) tell me what the pro's and cons are for having a character's parts (body parts, clothes and such) as separate meshes in a model or as one merged mesh? I think for game engines it's usually whole meshes but is it the same for animation? it strikes me handier for alterations to keep things separate but maybe that causes problems when you animate? Very curious what is considered best practice with this...
Edit: whoops seems a lot of extra questions arose while i was typing my post - i though it was resolved, sorry!
Can anyone (cough..like Eric..cough ;) ) tell me what the pro's and cons are for having a character's parts (body parts, clothes and such) as separate meshes in a model or as one merged mesh? I think for game engines it's usually whole meshes but is it the same for animation? it strikes me handier for alterations to keep things separate but maybe that causes problems when you animate? Very curious what is considered best practice with this...
Edit: whoops seems a lot of extra questions arose while i was typing my post - i though it was resolved, sorry!
-
- Moderator
- Posts: 754
- Joined: 25 Nov 2009, 01:41
- Contact:
Re: xsi 2013 merging envelope operators
@Kerro: In my experience I've tried to stay away from separate meshes as much as possible aside from clothes and accessories. The reason is that it is hard to keep the parts that join each different part consistently deforming properly. This is especially true in areas such as the shoulder where "skin" slides over bones and muscle mass and also where subdivisions usually help a bit for smoothing.
Since there is no geometry that is past the boundry edges on the separate meshes there is no way to get it to deform the same as if it was without a lot of extra work. I've never worked in games so I'm not sure about that part.
If you have a head that is separate that you rig and do your blend shapes on that is detached, it may be OK for you to keep it that way as long as there is some overlap between the shirt and the lower part of the neck. Otherwise its going to be the same issue as above.
With clothes and things like shoes you can usually use GATOR to get like 60% there if you're doing it via envelopes. You'll definitely have to setup some other rigging mechanisms that help slide things around or push in certain directions that is triggered by the foot roll or some other stable parameter. A good example of this is shoe laces. As they zig zag back and forth down the tongue of the shoe it will not be as simple as GATORing to the shoe or foot as the topology is very different and the weighting won't be consistent from foot to shoe and shoe to lace. This doesn't mean that merging them to 1 mesh is a good idea either. Just an example of something that you need to consider when extracting different parts.
You should also think about rig maintenance as well. Having tons of meshes to keep track of for weighting isn't terribly fun. I'm dealing with hundreds of meshes per rig on my current project. It can get a bit daunting but if you're organized, have tools to make simple tasks easy (mirroring weights from left mesh to right mesh), and there is solid reasoning for it then you're in good shape.
That's my take on it. The short of it is that you should keep the mesh as a combined mesh to keep deformations consistent over areas that deform a lot.
Since there is no geometry that is past the boundry edges on the separate meshes there is no way to get it to deform the same as if it was without a lot of extra work. I've never worked in games so I'm not sure about that part.
If you have a head that is separate that you rig and do your blend shapes on that is detached, it may be OK for you to keep it that way as long as there is some overlap between the shirt and the lower part of the neck. Otherwise its going to be the same issue as above.
With clothes and things like shoes you can usually use GATOR to get like 60% there if you're doing it via envelopes. You'll definitely have to setup some other rigging mechanisms that help slide things around or push in certain directions that is triggered by the foot roll or some other stable parameter. A good example of this is shoe laces. As they zig zag back and forth down the tongue of the shoe it will not be as simple as GATORing to the shoe or foot as the topology is very different and the weighting won't be consistent from foot to shoe and shoe to lace. This doesn't mean that merging them to 1 mesh is a good idea either. Just an example of something that you need to consider when extracting different parts.
You should also think about rig maintenance as well. Having tons of meshes to keep track of for weighting isn't terribly fun. I'm dealing with hundreds of meshes per rig on my current project. It can get a bit daunting but if you're organized, have tools to make simple tasks easy (mirroring weights from left mesh to right mesh), and there is solid reasoning for it then you're in good shape.
That's my take on it. The short of it is that you should keep the mesh as a combined mesh to keep deformations consistent over areas that deform a lot.
- Kerro Perro
- Posts: 306
- Joined: 20 Mar 2011, 01:01
- Location: Leiden, the Netherlands
- Contact:
Re: xsi 2013 merging envelope operators
Thanks for the insights Eric! I guess it also comes down to planning ahead about what parts deform "uniformly" and what parts have more unique movement like you described. I do wonder about things that basically maintain their shape like hats/helmets and shoulderpads, in a way these things only need to be constrained but i could also see a case for merging just to keep it organized and predictable...
Btw hundreds of meshes per rig? That sounds like a nightmare! ;) Also it sounds like a transformer
Btw hundreds of meshes per rig? That sounds like a nightmare! ;) Also it sounds like a transformer
Who is online
Users browsing this forum: No registered users and 26 guests