Isotropix Forums

Procedural nodes comprehensive introduction/overview

General Discussion about Isotropix and CG related topics

Procedural nodes comprehensive introduction/overview

Unread postby devinj898 » Tue Sep 01, 2020 12:47 pm

I recently saw this video about all the nodes in Blender
https://youtu.be/cQ0qtcSymDI
and I thought it would be a great idea if a video like this existed for all the nodes in Clarisse ifx’s node editor. I don’t know if there is anyone else out there that has been trying to learn how to use the nodes to create better, smarter, more powerful materials and other procedural effects in Clarisse, but I have been struggling with figuring out what all the different nodes do and can do. If this already exists then please send me a link. I think this is holding back my workflow and I’m not sure how to close the knowledge gap. I would think it could go something like explain how the node works, what it requires as inputs/outputs, what kind of data it outputs, and some basic use cases. Thanks!
devinj898
 
Posts: 25
Joined: Mon May 05, 2014 7:36 am

Re: Procedural nodes comprehensive introduction/overview

Unread postby dboude » Wed Sep 02, 2020 8:58 am

Hi,

A big majority of the nodes are self-explained, like all the math nodes, the color ones etc... There are other Clarisse specific nodes that are covered by our tutorials. If you have question about some nodes, feel free to ask.

Cheers
Démian
Isotropix
Technical Artist - Clarisse Specialist
User avatar
dboude
 
Posts: 1567
Joined: Mon Jul 03, 2017 10:51 am

Re: Procedural nodes comprehensive introduction/overview

Unread postby devinj898 » Wed Sep 30, 2020 3:44 pm

Maybe it is just me . . . When I saw this tutorial https://www.youtube.com/watch?v=9ngbOV0YKbk I was at first happy that there are more tutorials that get to the meat of what Clarisse can really do. But then I just started to realize how black-boxy the node editing in Clarisse is, overall and in the individual nodes; for instance, for the overall, the process of creating what is in that video, and for an individual node, quantize. I also realized that it is not just what the node does but when you need to use it and how it should be used for Clarisse's sake. There are more than a few times that I have been playing around with the nodes trying to learn/experiment/create something and at a certain point I just give up because they don't do what I am expecting them to do based on the Clarisse documentation that you refer to and general knowledge that I have about I am assuming the node does (like multiply multiplying two things). Like the image below, from the tutorial--I know what most of these nodes do, but Eric mentions that something isn't working how it should be, and so then he needs to add in another node to multiply/clamp/etc. the values, but as far as I'm aware there doesn't seem to be a way to understand why that is required unless someone just straight up tells you it has to be that way. A side-related question would be, how can you see what values you are getting in the output of a node other than to just hook up a color node and *try* to visualize it on an object. Again, it would be great to know if this is just me, or if anyone else has these issues too.

2020-09-30 10_31_19-Window.jpg
devinj898
 
Posts: 25
Joined: Mon May 05, 2014 7:36 am

Re: Procedural nodes comprehensive introduction/overview

Unread postby mdkai » Wed Sep 30, 2020 5:14 pm

Well I understand your point, as the manual is sometimes sparse with information. But the majority of nodes are indeed rather rudimentary and in such can be used easily with a lot of situations.

The most important thing to understand first, is that clarisse does not care for color, integer or vector conversion as in other applications. Data is used in a universal way similar to nuke, where three channels can be RGB or three different floats to drive different values. A grey value of 10 could be used for distance Z, 1000 could be blackbody shader data etc..

Once you grasp that any value could drive another you already understand.
As for what Eric mentioned is that the remap node by default clamps at 0 -1 because the main purpose is to remap data in that area ..But here again he uses wider values as they come from pworld node literally giving the pixel a world space value.

To preview any current selected node indeed there is not yet a method like in Arnold or Blender with nodewrangler. I assume this would be fixed in the next major release.

So in a way one could sum up all nodes and give a brief overview of their functionality but in what way to utilize them could take a while to explain as there are quite often many use cases...

Cheers
Kai
Join us at the Discord https://discord.gg/G5cJj9A
mdkai
 
Posts: 342
Joined: Tue Oct 07, 2014 7:24 pm


Return to General Discussion