Turn the page

[ikam]

Is it what I'm thinking about actually. I didn't choose yet how to manage materials.
wz4 modular material idea is nice and enough modular. or maybe use similary materials nodes editor (but with operators stack) like in unreal or blender, even more modular but I found it more difficult to use and need more time and effort to build a material, or prebuil-materials mixer but it is less modular but easier to use... what is certain is that there will be a code editor to manually write and edit shaders (like the custom material in wz4).
any idea ?

[CybeREX]

If it don't take enough time, I prefer two options.
Something like mega material operator (basic options for diffuse, normal map, specular, and shading options/alpha channels) mean all in one.
And modular material if we need create something advanced and complex.
PS. Just Idea

[Skinnytorus]

As for me, I like the idea of wz4 material. For what it's worth, building stacks is a modular approach as such. The problem is to ensure quick and flexible connectivity. Ability to drive any parameter by any other parameter is so tempting... . Anyway, guys... My idea is to stack operators with default connections but be able to drive operator parameters by any parameter in any operator connected to the stack. This is pretty much close to using global variables in wz4.

Implementing a node editor, being kinda trendy on its own, is not a bad idea, but thus we are breaking the idea of wz4 modularity. A much more feasible thing imho would be to find a way to group operators into compounds (groups) with ability to publish inputs and outputs, which then could be used as a preset.

[ikam]

Yes you're both right.

To resume multiple systems is probably the ideal solution :

- a single mega material for quick and basic design. A single operator that do the trick.
- a modular system to compose custom and advanced materials.
- and a manual code editor for programmers to write materials from scratch or just edit parts of a previous builded modular material an give full flexibility.

I already thought to a group operators system as presets and i like the idea, this is something I will probably work on (a bit like in houdini if you already test it) and not only for materials but for others operators family too.

Skinny : what I don't fully understand in your idea is : "be able to drive operator parameters by any parameter in any operator connected to the stack." Do you mean a a kind of convenient system where you can specify which parameter in which operator to export as global parameter so that it can be reused in a 'final' operator that imports all theses parameters to give the ability to drive them in a single place to avoid to go back in the operator tree ?

[Skinnytorus]

- a single mega material for quick and basic design. A single operator that do the trick.
- a modular system to compose custom and advanced materials.
- and a manual code editor for programmers to write materials from scratch or just edit parts of a previous builded modular material an give full flexibility.

Good. If it's not too much pain in the ass

I already thought to a group operators system as presets and i like the idea, this is something I will probably work on (a bit like in houdini if you already test it) and not only for materials but for others operators family too.

Awesome!

Skinny : what I don't fully understand in your idea is : "be able to drive operator parameters by any parameter in any operator connected to the stack." Do you mean a a kind of convenient system where you can specify which parameter in which operator to export as global parameter so that it can be reused in a 'final' operator that imports all theses parameters to give the ability to drive them in a single place to avoid to go back in the operator tree ?

Exactly. And it would be cool if the links were dynamic (driven values change during animation). Also, a good idea might be to make available such globals as free camera attributes (pos, angle, target, etc.).

[CybeREX]

just my 5 cent - I prefer more wz4/wz1 like style than connecting with arrows.

For me stacked operators looks better

[ikam]

yes It is was implied. All is done with the stack operator mechanism.

[CybeREX]

Little offtopic.
I just upgrade my old video card from Radeon HD 4650 512MB to Asus GeForce GTX 660 Ti DC2O-2GD5





Test render from my blender file on CPU was around 10 minutes (cycles render)
With using this new monster with CUDA in blender and switch to GPU hardware render same scene was 1 minute

(1280x720 resolution of the image render)

in 3d mark 06 points raised up from 4500 to around 20.000

I just happy now)

[Skinnytorus]

Good for you, man. Welcome to CUDA world I've been using nvidia cards since long long ago by now. And I love blender too. Can't wait until they optimize Cycles GPU volumetrics rendering routines. Too much noise this far...

[CybeREX]

I just find this short demo on pouet


I surprised, that it was made on unity
File Size are big for short one scene, but it's much smaller than in UDK stand alone player
Need to try unity engine)

And you can run demo direct from url website and go to full screen!
This is very nice feature for demo collectors site. Or portfolio, or demogroup site

[Skinnytorus]

Yeah. The demo is stylish, but it's just one scene and all the shaders seem to be written from scratch.
Imo, neither Unity nor UE is able to become a decent demo tool. It's just a different workflow.

[ikam]

I'm agree. I've played a bit with these engines (and others like) they are great and have pretty powerfull render and cool features but they not well designed for a demo workflow and are generally too heavy for simple stuff. In fact I never found any good alternative to wz and I don't think it exists (demo animation workflow, realtime 3d rendering, lightweight, highly customizable)... It's a bit why I deciced to code my own, there is no tool, so do it.

[Skinnytorus]

Any news on your project?

[ikam]

Still working on the render engine.
and finished the wz4 plugin to use wz4 bitmaps and meshes.

[Skinnytorus]

Good. I just can't wait for the time the proto becomes available )