Jump to content

When is a network too complex?


art3mis

Recommended Posts

Many of the solutions I've seen here in the forums are absolutely brilliant and reflect a great amount of hard work as well as experience.

But many posted examples are also incredibly complex and almost impossible to modify, debug, and art direct.

Is there a point when you are working on a solution for a particular effect when you have to say,' Wait a minute, I'm creating way too many nodes at the SOP level' and start over again, trying to simplify?

Or do some solutions simply require 100's of nodes (and or tons of Vex code)  to achieve the desired result?

Edited by art3mis
Link to comment
Share on other sites

Some node networks can be very complex like hundreds or even 1,000+ nodes. When networks get that complex it's important to stay organized using network boxes and subnetworks. Documenting things and leaving notes can make or break a job especially if there are others touching it. Simple is better but that's not always possible. Hopefully it goes without saying but don't do something with a dozen nodes that could be done with one or two nodes.

  • Like 1
Link to comment
Share on other sites

I've often looked back on the network(s) I'm using and thought "shit - this is crazy" and tried to go back and simplify - it's a good idea generally

but it's a mistake to look at Houdini's node network(s) and think there are 'too many nodes'

every DDC app worth using builds the same insanely complex networks - the difference is Houdini shows them to you...

hypergraph.jpg

Link to comment
Share on other sites

When it does not serve your objective.

Im not a big coder, but I once heard a programmer say it is important to not add features freely. Every node cost efficiency. And there is a universal wisdom of "using the least effort".

I find it easier to understand complex networks that you have built then those built by others. So don't think you don't have the capacity to work on that level.

Edited by Yon
Link to comment
Share on other sites

maybe when there are inter dependencies within in it that would be better separated. which makes things easier to change.

say for example a dop network containing related flip, pop and smoke solvers, it seems to make sense to run them as three separate networks so if there is an issue/change with one part of the network you only have to re run that dop.

 

  • Like 1
Link to comment
Share on other sites

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.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...