I'm just getting started with houdini, but i can't wrap my head around the fact that many different operators in houdini require different networks. What is the advantage of having it this way? For example, in the ideal world (in my head), a project consists out of only one network, where my lights are stored, the camera's, the objects and object modifiers, the particles, the materials, etc. Why is there a need of having different types? I can imagine having thousands of tools at hand could be a reason, but this could be nicely sorted in different tabs (like touchdesigner)
There will most likely be a very good reason which will trigger me to do a facepalm, but until then, i'm very curious
It's an old discussion - there would be an enormous amount of nodes in a flat context. We've never had confirmation from SideFx if it's a technical limitation or more just keeping the old and common user experience or whether its simply too much work to merge everything for an unknown reward, or even rejection from the user base.
the change, although probably beneficial, would have to go against “if it's not broken, don't fix it” approach. and it would involve a lot of development time if not an overhaul. one way to avoid the mind boggling amount of nodes in a complex scene is to have vertical split contexts within an object, with options to toggle on/off contexts, show relevant connections from one context to another, etc. bird's-eye view of the scene, with options to focus in or out the context you're working with (new node automatically created and placed in the proper context). that's what I'd like to see, given that the trend of houdini development is to consider an object as an asset/scene assembly/hda/etc.
Connections of different contexts represent different data types. A SOP has geometry type connections. COPs have image type connections. An object has transform connections, but also an implied geometry connection that is always the displayed node inside. A flat context might allow interesting things like wiring geometry to multiple object's geometry connection or computing a transform using sops and connecting to an object's transform, but that would lose a lot of the simplicity of not having to worry about types when making connections.
Okay so at the moment i have spent a whole bunch of days/weeks of learning how to work with houdini from documentation, tutorials, etc.
Then i saw a few tutorials about solaris (since it's quite ‘'new’', it wasn't really represented that much in older tutorials)
I still don't really get it. It seems that solaris replaces the OBJ-level when it comes to lighting. but not just that. also instancing, scene building, shading, etc. It seems that you can also create geometry (sopcreate), edit it, and create pop networks inside of those sop=networks.
So i guess my questiom is: Does solaris/stage/LOP (they are practically the same?) replace the OBJ level completely? Why would i want to work in the default obj level when i can do all of the creation inside of a solaris/lop network? what is the benefit? If i create my objects in the OBJ level, i have to import the whole scene again in solaris, which takes a bunch of extra, unnecessary steps?
The more tutorials i watch, the more confused i get. It feels like i'm in the middle of a philosophy-shift of houdini.
/obj/ gives you the advantage of a lot of existing tools, tutorials and so on. Rigging comes to mind as well.
/stage/ gives you lots of juicy new tools like the lightmixer etc. It is not very Houdini-esque though. Yes, you have nodes, but you always have to keep in mind your “Scene Hierarchy” next to your node hierarchy which you will get to love in SOPs.
