ICE repeat

Discussions about SOFTIMAGEs© Interactive Creative Environment©
Post Reply
wesserbro
Posts: 177
Joined: 27 Oct 2012, 18:05

ICE repeat

Post by wesserbro » 08 Nov 2016, 14:10

As i understand, 'repeat' node repeats ice-tree on a point N times, then ice goes to the next point and repeat the tree N times, right?
But can i force it to run on each point once, then on each point again second time and so on, but within a modeling (NOT SIMULATED) ice tree?

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: ICE repeat

Post by Mathaeus » 08 Nov 2016, 14:41

Yeah you can use it in non simulated ICE tree, in almost the same way as Houdini's for each sop or vop. Exactly it repeats everything plugged in certain execute node, so, not entire ICE tree. That is, you can use with something like get my_value > do some accumulated calculation > set my_value. In next ICE 'row of nodes' you can do it again.
Some typical use is smoothing of some attribute according to neighbors, point position, weight map value or such.
However, it's good idea to avoid the ''repeat'' node if possible, these iterations are significantly slower.

wesserbro
Posts: 177
Joined: 27 Oct 2012, 18:05

Re: ICE repeat

Post by wesserbro » 08 Nov 2016, 15:41

Thanks for your reply, Mathaeus!
But while dealing with iterations, i've noticed another problem with my tree ) So to not create a new topic, maybe i'll describe it here.
What I'm trying to do is conform subdivision object to another by having 3 meshes: original lowpoly(sphere, red), mesh with subdivision operator from that lowpoly (polymesh, green), and a copy of lowpoly (sphere1, blue). So I subtract position of subdivided vertex from vertex position of a object to conform to. Then i add this difference to vertex of original cage so that subdivided vertex become closer to desired position. within several iterations each vertex of subdivided geometry(polymesh) should lie or be very close to sphere1.
But I cant get the real position of vertices of subdivided mesh, it always gives to ICE the same numbers. So the question is can we get real points position of subdivided mesh(polymesh) after its lowpoly cage (sphere) was modified by ICE on previous iteration?
si.gif
As we can see from this picture, left outside vertices of the green mesh visually (perspective view) have greater X value than the correspondent vertices of the blue mesh, but ICE 'value' boxes show us the opposite.

wesserbro
Posts: 177
Joined: 27 Oct 2012, 18:05

Re: ICE repeat

Post by wesserbro » 08 Nov 2016, 16:31

If someone is interested, here what I get manually repeating freezing and applying my ICE compound several times:
SubDtest.gif
top - basically what i wanted ) except automatically iterating in ICE
bottom - regular subdivision
Last edited by wesserbro on 08 Nov 2016, 16:54, edited 1 time in total.

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: ICE repeat

Post by Mathaeus » 08 Nov 2016, 16:37

wesserbro wrote:Thanks for your reply, Mathaeus!

But I cant get the real position of vertices of subdivided mesh, it always gives to ICE the same numbers. So the question is can we get real points position of subdivided mesh(polymesh) after its lowpoly cage (sphere) was modified by ICE on previous iteration?
I'm afraid this is dependency cycle. I guess you're using modeling operator (model panel > create > subdivision). Perhaps it's possible to combine something with duplicates, anyway should be much easier if subdivision is calculated directly by ICE.

wesserbro
Posts: 177
Joined: 27 Oct 2012, 18:05

Re: ICE repeat

Post by wesserbro » 08 Nov 2016, 19:50

simulation so far (using 'subdivide topology' node on another mesh)
Image

Post Reply

Who is online

Users browsing this forum: No registered users and 40 guests