panel system test

[plastic_]

I've been sporadically developing something for about 1 year, I thought it is time to resurrect it.
re:FORM is also developing a similar solution as it seems, I watched the thread with interest, I'm curious how it will turn out. (There are many ways to approach it in max)

My aim is getting some functionality of Rhinos great PanelingTools into 3ds max:

http://en.wiki.mcneel.com/default.as...gExamples.html
http://ncodon.wordpress.com/tag/panelingtools/

It looks very promising, performance wise. The problem/complexity lies in the initial calculation, interpolation between face normals, partial rebuilding of changed topology, etc, etc.
This is where I'm not done yet. I'm going to vacation, afterwards I will work on it. Estimated release in september.

What I can say is that it will require a poly base mesh with quads, as distribution surface. The panel elements can be instanced (but don't have to be). Multiple panels are possible, distributed by random, material ID, etc...

Apologies for the video capture, it turned out too long, and the screen capture software was eating my quad menus...but you can skip to the end for the best parts with VRay RT glory

==============
ENOUGH TALK HERE IS THE VIDEO: http://www.marclorenz.com/video-tuto...nel-system.htm

(no interface yet, only proof of concept)



==============

[Sawyer]

I am interested in where this goes. Good work!

[RE:FORM_STUDIOS]

Wow, you've managed to put my coding skills to shame. How have you managed to get the realtime feedback on the mesh? Is the script written in the SDK?

Are you applying the panels to the base mesh and then simply adding the distortion modifiers onto the panels? Or are the panels correctly deformed based on vertex normals?

Looks great! and most of all FAST!

[dellis]

That's impressive! I'll be keeping an eye on this!

[peterguthrie]

looks great!

let me know if you need someone to test it

same goes to you Patrick!

[Neilg]

Also throwing my hand in as a potential tester
Looks great, very suprised how quick the feeback seems


Whats the plan for the interface (options/etc)?

[plastic_]

Are you applying the panels to the base mesh and then simply adding the distortion modifiers onto the panels? Or are the panels correctly deformed based on vertex normals?

yes and yes.
Each panel is a node and deforming itself to the face it was assigned to.
There is some effective but also nasty hack involved I can't talk too much about right now, until I sorted things out

I would love to do it in C++, that would make everything realtime, even the initial calculation. But I would have to get help there with the coding. Compiled plugins are also a support nightmare, with recompiles needed for each new max version, couldn't do something like that for free. I think with a commercial plugin the demand might be too low to justify the costs. It's a tough call.
Most likely it will remain simple and free. I guess if this things are getting more popular, the VrayScatter etc. guys will include that functionality just along the way in like 1 day.

Also throwing my hand in as a potential tester
Looks great, very suprised how quick the feeback seems

Whats the plan for the interface (options/etc)?

No plan yet, I have to think about it. At most basic level "pick surface and "pick panel".
Some viewport degradation management will be important. I had some stability issues when "cached D3DX meshes" was enabled and I changed the topology of 1000 panels at the same time, while max was redrawing them in the viewport.

I'd also love to support Vray proxies, but unfortunately they don't like being deformed.