Ben's awesome patterns issues.

   1611   2   1
User Avatar
Member
408 posts
Joined: June 2015
Offline
I'm actively trying to learn as much as i can about drawing patterns with points or animating particles following specific patterns, and consequently found “Ben's pattern” hip file very interesting. I don't know this gentleman personally, not even his last name, so for credits follow the link please. (https://vimeo.com/244035351).

First issue. I've been trying to implement this file in which all the work takes place in a SOP solver using points created by a scatter node, tried with the humble means i have to replace that scatter node by a POPnet but somehow the SOP solver only picks up the point emitted before the “start frame” of that solver and ignores all points created after that. Why please?

Second issue. I'm trying to create curves from those patterns, or trails, or just connecting the points, but using the add node with the id attribute doesn't work. apparently all the points get strange “-1” id values, and i have no clue why. Of course it isn't a simple file to look into as Ben's coding isn't meant for noobs like me, and if i can make sense of some bits, most of it still flies way above.

Thanks in advance.
Edited by Adriano - April 7, 2018 17:01:24

Attachments:
patterns2.hiplc (668.0 KB)

User Avatar
Member
36 posts
Joined: May 2015
Offline
Oh boy, quite tricky indeed. Basically the problem seems to be retaining an attribute that acts as i@id would but for each independent “set” of patterns. So fare none of the attributes existing in those points work with that logic.
I've tried to find a way to let them inherit the id from the input geo but, as that one is constantly changing, I couldn't.

I am afraid I am a bit lost right now, probably a bit more digging could help, but for now Ben's method is not giving you any way to connect points with an attribute out of the box.

Hopefully somebody more skilled than me in coding will jump in and solve this one!
User Avatar
Member
408 posts
Joined: June 2015
Offline
Cristotek
Oh boy, quite tricky indeed. Basically the problem seems to be retaining an attribute that acts as i@id would but for each independent “set” of patterns. So fare none of the attributes existing in those points work with that logic.
I've tried to find a way to let them inherit the id from the input geo but, as that one is constantly changing, I couldn't.

I am afraid I am a bit lost right now, probably a bit more digging could help, but for now Ben's method is not giving you any way to connect points with an attribute out of the box.

Hopefully somebody more skilled than me in coding will jump in and solve this one!

Thanks for looking into it, man. Much appreciated.

This is how far we went, and by “we” i should say Johan Boekhoven (http://johanboekhoven.com) who's been kind enough to work a bit of his magic to patch things up on top of Ben's code. He managed to connect the dots in quite an ordered fashion, but the set up seems to break when the “random” split feature of the pattern is turned off. The “StartPoint” of each branch seems to jump position, along with the Ids changing quite erratically along the process. I'm convinced he could figure out a way, but i think now his mind is set to just rebuild this from scratch with POPs to enable more features in terms of pattern options and overal control on the points. We'll have to wait until he has time for it, Hopefully i can sponsor him a bit as well to push this set up to the extremes and make it a useful assets for everyone. It has such cool potential, but somewhat comes a bit short in its current state as it's way too hard to mesh it properly. I'd still love for short term applicaiton to fix that last little bit though so if anyone's indeed is up for the little challenge please jump in. Here is the current file.

Cheers,

A.
Edited by Adriano - April 9, 2018 16:09:55

Attachments:
Patterns_Polylines.hiplc (5.3 MB)

  • Quick Links