anthonymcgrath Posted November 20, 2016 Share Posted November 20, 2016 hi guys so i'm doing a simple test - a simple poly sphere setup with auto fracture hitting the groundplane. I export this as a .abc and it imports into maya easily enough however i need a bit more than that: currently the .abc comes in as one node - i need it to come in as a group of meshes. if possible i need the transform of each mesh to be centred to that chunk/piece/fragment and finally (this one is a big ask i know) is it possible to get the keyframes from here to maya too - basic translate/rotate/scale and visibility ? I know this is a big ask i'm really struggling with this one - i tried connectivity and partition sops after my simulated mesh results but it doesn't seem to work anyone got any ideas please? You'd be helping me out MASSIVELY if so! I've attached a super simple sphere auto fracture scene file for convenience thanks! ant autFrac_ball_example.hip Quote Link to comment Share on other sites More sharing options...
catchyid Posted November 20, 2016 Share Posted November 20, 2016 (edited) Hi, inside the sphere node, if you display the geometry attributes (pls, see attached), you would find name primitive attribute, which is basically unique for each piece or chunk. You could use "For each" node to iterate over these pieces using name attribute and store each mesh individually (if you don't know how to use for each node, I remember there is a master class that explains it, if you cannot find it, let me know and I will look for it). I "think" when you save a sequence of alembic files you already save position/rotation, so this done automatically, actually I remember I saved once a sequence of animation as alembic file and import in Maya and I got the sequence position/rotation correctly. As per changing the chunk pivot, I don't know but actually, I "think" that alembic (and I am just guessing here) would save a set of primitives in world space, i.e. you won't need a pivot to move chunks as in each frame the entire primitives would move in the world space (again, just guessing ) Edited November 20, 2016 by catchyid 1 Quote Link to comment Share on other sites More sharing options...
anthonymcgrath Posted November 20, 2016 Author Share Posted November 20, 2016 that is an awesome explanation thanks @catchyid Yep i think i'm gonna have to watch that for-each tut again i know the one your on about I did some googling and some chap did mentionusing a for-each and using a delete node and then delete non selected then merging back in - i'll have a look into tackling it that way possibly. Quote Link to comment Share on other sites More sharing options...
catchyid Posted November 20, 2016 Share Posted November 20, 2016 good but most probably you won't need delete node as for each would select a chunk as a whole, i.e. in each iteration would be given a complete chunk, so all what you need to do is save it...good luck 1 Quote Link to comment Share on other sites More sharing options...
anthonymcgrath Posted November 20, 2016 Author Share Posted November 20, 2016 interestingly doing the same test with a pre-fractured object (usual trick... isoOffset + scatter) gives me separated pieces. it may be that the assemble sop further up the chain is helping separate out those pieces ? to add i have a slightly different setup from my scene file as well where i'm using the transform pieces node to copy my sim from low res geo to hi res version that involves using packed geometry with the assemble node and then using a dop import but using it to bring points in and then transferrring to the hi res geo. https://www.sidefx.com/forum/topic/26806/?page=1#post-171070 Quote Link to comment Share on other sites More sharing options...
anthonymcgrath Posted November 20, 2016 Author Share Posted November 20, 2016 4 minutes ago, catchyid said: good but most probably you won't need delete node as for each would select a chunk as a whole, i.e. in each iteration would be given a complete chunk, so all what you need to do is save it...good luck cool thanks - and thanks for the pic and explanation Quote Link to comment Share on other sites More sharing options...
catchyid Posted November 20, 2016 Share Posted November 20, 2016 I gotta go quickly but just a quick note: I think it's better to always prefracture your mesh before going into simulation, it gives you lots of control of how the fracture should look like, once you put assemble node you can as well control many physical parameters of each piece (like mass, bounce, ....), also you can build constraints networks which you can control easily to determine exactly how your destruction shall occur... there is a master class that explains that in details, just search for bullet and it will be named something like Bullet Master Class H13. 1 Quote Link to comment Share on other sites More sharing options...
anthonymcgrath Posted November 20, 2016 Author Share Posted November 20, 2016 Excellent thanks I've setup a system at the moment where I'm transferring low res to hi via a transform geometry setup. The low res has the assemble sop pre dops etc.. when i bring that high res into my scene in maya it is all separated. As you say the pivot sits at global zero per chunk but there has to be some way to fetch that over centred up per piece and potentially with keyframes or that may need to be something I script maya side really. On a side note.. I did an auto fracture earlier mainly because I wanted to see bigger chunks get broken into smaller ones and I don't know how to do that via a voronoi/scatter setup really. I would need to have thousands of tiny pieces clustered together and that gets super heavy from what I can see. Is there a way to have secondary fractures on generated pieces that have been simmed? So they essentially do a secondary fracture. Quote Link to comment Share on other sites More sharing options...
catchyid Posted November 22, 2016 Share Posted November 22, 2016 Sorry... I did not check my email yesterday As per your second question: I've watched couple tutorials and asked questions on this forum, also read Houdini docs, from my understanding is that auto-fracture is just there to make quick shots, and according the documentation, it's good to be used for background objects where you don't need much details. It's possible and actually easy to control the shape of your fracture using Voronoi fracture, just place highly dense points cloud on places where you need lots of pieces, for example: if a missile is to hit a building, then place a dense points cloud on the impact area and a minimal number of points around the building... note that you would create RBD Packed Glued object which means you would have a constraint network where you can control the strength of the bonds, if you change that strength parameter, you will see different types of gradual destruction (i.e. very strong glue bond, the entire object will act as one piece, a weak strength will make all pieces collapses at once...). You can also break the bonds procedurally in the DOP network, if you need to. I would highly suggest that you watch Bullet Houdini 13 Master class, it has all these techniques explained in details As per your first point, well, I really don't know I think, it's best if you simplify the problem, just have a sphere some where in the scene and write it alembic and open it in Maya and see how its transformation is copied, there might be some options, somewhere, ...Note that both Maya and Houdini concatenate transformations, they use the same concepts (which are universal in any 3D app), so you just need to know how each application handles its transformation in order to transform between them... Good luck Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.