[Plugin] SketchyFFD (Classic)

REGIS

On simple shapes it's ok but on more complex it's not the same.
It seems to do triangulate faces on coplanar quadrangulate faces

FFD subdivide option is great on cube with square angle because he cuts the cube on RGB axes.
For those how know blender there is an option in the k key : select a face an select loop cut.
I would like to have the same option in sketchup.

justry

What Dale's plugin does is called a catmull-clark subdivision. Google it and you'll find lots of pages explaining the algorithm. It's not that hard to understand once you've went through it a couple of times.

oecodesignator

Hello Chris.

Applications? Last night, I was talking to our design manager and he was looking for a capability to add quickly in my landscape architecture model a twisted (literally, not the sicko type, hehe) building. I failed to get a copy of the script previously but read Catch Up had previously enough to know it could be done. I'm downloading the script and at the same time posting this to say, "Yes, UAE has been waiting for a script like this to keep the rapid-paced development moving. Cheers!"

and thanks.

Rico

@cphillips said:

Has anyone actually used this script for anything useful yet?

Chris

CPhillips

Thanks Rico!

mikaztro

Where can I download the latest FDD plugin. Thanks.

Gaieus

See Chris' post when he updated the link:
http://www.sketchucation.com/forums/scf/viewtopic.php?p=60071#p60071

mikaztro

@gaieus said:

See Chris' post when he updated the link:
http://www.sketchucation.com/forums/scf/viewtopic.php?p=60071#p60071

Thanks mate!

pav_3j

this is brill, thank you so much

pav

svetik

Thanks man-it is really inportant emproovement for SU. Google is too poor to do it
BUT! it does not work for my favorit 3D soft- AtLast Sketch Up 5 Why?

remus

This is a bit of a long shot (aka a feature request) but would it be possible to make it so that grouped geometry (within the group to be FFDd) could be moved by the control points but not deformed. In an ideal world the edges that were 'shared' between the group and the geometry outside the group would also remain unaffected.

I was just thinking that if this worked you could use FFD a a very basic method for manipulating rigged meshes, perhaps in combination with monsterzeros rigging hack.

CPhillips

@remus said:

This is a bit of a long shot (aka a feature request) but would it be possible to make it so that grouped geometry (within the group to be FFDd) could be moved by the control points but not deformed. In an ideal world the edges that were 'shared' between the group and the geometry outside the group would also remain unaffected.

I was just thinking that if this worked you could use FFD a a very basic method for manipulating rigged meshes, perhaps in combination with monsterzeros rigging hack.

I dont think I follow. Can you post a picture?

CPhillips

@svetik said:

Thanks man-it is really inportant emproovement for SU. Google is too poor to do it
BUT! it does not work for my favorit 3D soft- AtLast Sketch Up 5 Why?

I think it should. But I dont have SU5. Can you open the ruby console try it and see if it prints any errors?

topmanblue

thank YOU sir!!!

remus

Imagine we start with this mesh:

The box is a group within the group to which FFD has been applied.

Currently deforming this mesh does this:

What i was suggesting is to get it to do this instead:

CPhillips

I don't follow. If you want to bend the boxes like that why dont you do them separatly.

remus

The two bits that have been affeced arent important. What im trying to describe is the way FFD treats groups within groups.

If you look at the box, notice it is left behind in the first instance, but affected in the second. thats what im trying to suggest.

CPhillips

Ah. Dealing with subgroups gets a bit tricky, but I will see what I can do.

remus

I look forward to seeing what you come up with

plot-paris

a very interesting idea, remus.

but what happens, if the deformation is a bit more complicated?

imagine you want to alter the shape of the first object (on the left hand side of the image). the red plane is a group within the group of the green bowl.

if FFD behaves like in your sample, it would look like the second form - the bowl is deformed and the red plane has to decide, where to stick to the connecting geometry.

or should the group within the deformation object be deformated where it is necessary in order to keep contact to the connecting geometry at every common edge?
but would that make any difference to simply enabeling FFD to alter nested groups no matter how deep the tree structure goes?
it probably would, because the group's content would only be changed, when absolutely necessary.

I am confused now...

remus

I was thinking of it mainly in the context of using it to rig models, so i didnt really think about situations such as the one you mentioned.

Basically ive got no idea how it would behave in those sort of situations, i suppose thats up to chris' discretion, if the idea works at all.