Hi Tim, Not sure if this helps your case, but I posted a few complied array nodes on the list in these threads: http://bit.ly/TpZH22 http://bit.ly/1qQe0v4
On Thu, Jun 26, 2014 at 3:36 PM, Tim Bolland <[email protected]> wrote: > [CHECK DROPBOX LINK FOR IMAGES] > https://www.dropbox.com/sh/wkk5r0bl9cisx5p/AAD7pSEDA4pVb2XItmqKD8S1a#/ > > > Hi everyone, I have a little puzzle I'm hoping someone can help me with. > I'll start off with what I'm trying to achieve, followed by how I'm trying > to do it and where I'm getting stuck. That way if you think my approach is > crazy and I need to be slapped with a logic stick then you can let me know! > > I'm looking to blend between a bunch of objects that start close together > but will pull apart, as they pull apart we see that there are loads strands > between them, and that large parts of what we thought was solid was > actually made of fibers. > > Because there will be a few objects I will need a robust system that lets > me define the direction the fibers 'flow'). Curves seems a good choice here > > - As a simplified test I took two grids and extracted curves from the > edges (I can later extract curves from more complex shapes). > > - I emit points evenly along one of the curves and query the curve(s) > nearby. > > - I then create what are essentially strands between them, although I'm > only concerned about the point positions of these 'strands'. > > - I then generate stands flowing in the direction of the curve between > each the previous strands, either blending linearly or using a bezier curve > interperlation. (see picture to see what I mean) > > It's this final step I'm getting stuck with, If I take these strand > positions and feed them into a fit bezier curve then a I end up flowing the > wrong direction i.e along the original strand. It appears I need to do > some serious array wrangling to get the position array in the form I need. > As you can see from one of the pictures I managed to get it sort of working > using a select in array node, but I'm having to manually build this array > from scratch for the stand positions to join completely. Ideally I would do > this without a repeat node. > > My explanation here is probably very bad, so check out the images to get > an idea of what I'm doing. This might also be useful to see where some of > my inspiration came from: > > https://vimeo.com/11079196 > > Regards, > > Tim > --
