Group layers are unique!

synthsin75 · Post by **synthsin75** » Sat Apr 26, 2008 11:31 pm

Some of the pros probably already know this, but I just found out that group layers aren't just the retarded cousin of bone layers. From what I had read, I had assumed bone layers did all the same stuff with the addition of bones. But not so.

The one feature unique to group layers is that they preserve the layer rotations of their child layers. Bone layers collapse any rotated layers to their single plane.

Just an aha moment I thought worth sharing.

Rasheed · Post by **Rasheed** » Sun Apr 27, 2008 7:53 am

I don't understand what you mean. Could you give us an example, please?

heyvern · Post by **heyvern** » Sun Apr 27, 2008 8:11 am

I'm not 100% sure but this seems like an odd new bug.

If you "3D" rotate vector layers inside a bone layer the layers don't actually "rotate", the points "squish" and stretch to mimic the shape that would be produced from layer rotation. It is very very strange.

I don't recall seeing this behavior before. I could swear I have rotated vector layers like this inside a bone layer... Could this be something new introduced with the last update or am I going crazy?

Was it like this all along and I missed it?

-vern

synthsin75 · Post by **synthsin75** » Sun Apr 27, 2008 8:18 am

http://www.mediafire.com/?xhaymnjmjjx

This file shows the difference. I have two vector layers, one of them rotated 90 degrees. I duplicated these and put one set in a group layer and one in a bone layer. At frame zero, they are identical, but on any other frame the bone layer collapses the rotated layer to its plane.

Use the orbit view tool to see what it is actually doing.

This means that group layers actually do have their own unique functionality.

P.s. I just found out that the bone layers only collapse the points to their plane, but not the vector lines.

synthsin75 · Post by **synthsin75** » Sun Apr 27, 2008 8:22 am

Vern, I think it has always been like this. I remember being puzzled by bone layers doing this early in my AS experience, but at the time had no reason to question it. The only reason I found it now was that I was think of how to best impliment your master bone script for controlling something like Rasheed's 3d walk rig.

heyvern · Post by **heyvern** » Sun Apr 27, 2008 8:23 am

Okay, just checked this in Moho. It has always been like this. How odd that I never encountered this before.

I do seem to vaguely recall a discussion about this possibly... can't remember the details.

-vern

Genete · Post by **Genete** » Sun Apr 27, 2008 9:09 am

Select the points at layer rotated 90 degrees. Then go to menu->Bones->Release points. Then bone layer works the same than group layer

It seems that although you don't have any bone inside the bone layer the influence of "not being at frame 0" produces a projection of the non released points to the skeleton working plane. I guess that all bone operations are done using 2D matrices so all the Z information of the points is automatically deprecated.

Rasheed · Post by **Rasheed** » Sun Apr 27, 2008 1:49 pm

Another possibility to fix this "bug" than to Release Points is to Flexi-bind the vector layer via the Bone menu.

I think this "bug" can only be solved with three dimensional bones. I think it shows which compromise had to be made to simplify bone layers.

synthsin75 · Post by **synthsin75** » Sun Apr 27, 2008 6:44 pm

Hey thanks Genete and Rasheed. Um.....so what are group layers good for again??

Oh wow, I was just messing around with this. If you bind that rotated layer to a bone, it still tries to move on the plane of its layer. This means that it will only translate by bone on one axis. Granted, it's stil collapsed, but may be useful for something.