nullCOMP, nesting, hierarchy MESS

Hello world )))
Probably I miss-concept here something in TD:

I need to have a rectangle cube (actually it’s a lower part of a virtual cake, then I will repeat with more pieces/higher stages)

So:

  1. I made one side of it, from a gridSOP geometry/texture/lights(rotating!)/cam and some effects on texture.
  2. I put it in nullCOMP
  3. I copy/paste 4 nullCOMPs and rotate them 90/180/270 degrees, so each became a side of a rectangle cube
  4. Then I put them all in another nullCOMP and rotated it

Questions/Problems:

  1. While geometry and cameras are doing well inside nullCOMPs lights are not - they doing well only in first (original) nullCOMP (even I’m not sure), in all copies I don’t understand what they rotate around (do they have kind of global coordinates?)
  2. Probably I miss-concept nullCOMPs nesting and have to link them (they have ins/out on top/bottom) instead of putting one in another? I was making it couple of years ago but do not remember now…
  3. How to put one cam/render on all 4 nullCOMPs, probably have to put render passes? How to link them from different nullCOMPs to one camera/render?
  4. Is there is a way to assemble all of it in one COMP/container? But how to rotate them all together in front of one camera keeping (in this case) effects/lights rotation around separate objects?
  5. Probably it’s completely wrong approach? :unamused:

attached:
cube_test*.toe - the situation I’m describing
cube_test_TDforum_01.1.toe (16.5 KB)
singleside*.toe - the initial setup, working good
singleside.1.toe (7.85 KB)
unwrap-helper.jpg - texture in case it got lost from movieinTOP

I’m not that new in TD but this is a new field for me… also hoppe it would be helpfull for others!
For texture 3Ding I used this tut by Ian Shelanskey: youtube.com/watch?v=P2PRFEAPpIU

thanx in advance
Andrey

EDIT

Looking more closely at this, I think part of what’s happening is that you’re applying series of transformations - one relative to the circle path around the circle, and then another as a rotation around the centroid of the top most null.

With that in mind the solution I came up with here was to use a CHOP to to extract the position data, then use that to target the transforms for the lights. A simple lookup network streamlines that process.

A render pass set up helps clean up the rendering for unique objects and lights per component.

Finally a little selecting and container organization lines up your outputs left to right.

Woof. Good challenge here.

cube_test_TDforum_01.mr_edits.toe (19.2 KB)