[Plugin] SketchyFFD (Classic)

kwistenbiebel · 8 Feb 2008, 02:58

...exactly the same here. I see no errors in the ruby panel though.

CPhillips · 8 Feb 2008, 03:21

...exactly the same here. I see no errors in the ruby panel though.

Wow. I tried it on 2 different machines and it works. Maybe you are using a Mac? Altho I don't see why that would make any difference..

kwistenbiebel · 8 Feb 2008, 04:38

Does the plugin in any means uses web based 'action script' or 'javascript' as I had trouble before with corrupt Iexplorer settings?
(I refer to web panel based plugins like Skindigo that failed in the past on my pc due to 'script blockers' )

boofredlay · 8 Feb 2008, 04:49

It is working fine on my home machine as well as my work machine. Both pc on xp. One new, the other 8 years old.

CPhillips · 8 Feb 2008, 04:51

@kwistenbiebel said:

Does the plugin in any means uses web based 'action script' or 'javascript' as I had trouble before with corrupt Iexplorer settings?
(I refer to web panel based plugins like Skindigo that failed in the past on my pc due to 'script blockers' )

No other scripts at all. Just pure ruby.

Hrmm. Very strange. Try doing a 2x2 FFD on a simple box. With the ruby console open in case of errors.

Dave R · 8 Feb 2008, 13:35

Very cool script. Thank you.

I do seem to being doing something wrong with it though. (Leave it to me to break it. ) See the attached screen grab.

I drew a sphere centered on the blue axis and above the ground plane. After moving a single point marked at the top and updating, the egg moved off to where it is now. What did I do wrong?

krash88 · 8 Feb 2008, 13:39

@cphillips said:

@krash88 said:

It looks great in it's current form, but I get this error in the console... Maybe I'm doing something wrong...

Error: #<NoMethodError: undefined method definition' for #<Sketchup::Group:0xbec0960>> C:/Program Files/Google/Google SketchUp 6/Plugins/SketchyFFD.rb:224:in analizeLattice'
C:/Program Files/Google/Google SketchUp 6/Plugins/SketchyFFD.rb:60:in updateFFD' C:/Program Files/Google/Google SketchUp 6/Plugins/SketchyFFD.rb:13 C:/Program Files/Google/Google SketchUp 6/Plugins/SketchyFFD.rb:13:in call'

Ooops! I fixed it and updated the script in the original post. Kwistenbiebel this is probably the same bug you were seeing.

Chris

It works great now! thanks for the fix!

CPhillips · 8 Feb 2008, 18:14

@dave r said:

Very cool script. Thank you.

I do seem to being doing something wrong with it though. (Leave it to me to break it. ) See the attached screen grab.

I drew a sphere centered on the blue axis and above the ground plane. After moving a single point marked at the top and updating, the egg moved off to where it is now. What did I do wrong?
[attachment=0:wa4rvigs]FFD1.JPG[/attachment:wa4rvigs]

It looks like you did a "Update FFD" without first closing the control point group.

Chris

Dave R · 8 Feb 2008, 20:48

Yup. That's what I did alright.

Thanks for the help.

Dave

Dave R · 10 Feb 2008, 23:07

Coen's post makes me ask, is there anyway to make this work on a component instead of or in addition to a group?

Dave R · 11 Feb 2008, 11:22

Where's the "dope slap" smilie?

I was doing it in reverse, trying to make a group of a components. Thanks Coen. I gotta think inside the box.

CPhillips · 11 Feb 2008, 21:18

@unknownuser said:

You're welcome Dave!

What's a dope slap though?

Also, I have been experimenting more with this script. I had an idea to position the cloned Construction Points along a series of curved Edges that you usually see used on Soap Skin Bubble examples, but I haven't been successful. I am trying to see (and would like Chris to shed some light on this) if the Construction Points have coordinates assigned to them. What I mean is that if I clone the Construction Points from the top and use them in the bottom will this distort the FFD? Should I therefor use the top ones? Well I tested this and it does seem so that there are coordinates assigned to them because by running two tests, one with using the Construction Points from above and the other test by using the bottom ones, and the result were different. The second test is what you see below, which brings me to another point - it is interesting to see what kind of curvy geometry you can create with this script, cloning Construction Points and positioning them. I hope I can find a way to understand how to exactly control them so it becomes easy to make any kind of curved geometry I had hoped to create in the first place.

As it is currently written it creates the construction points at fixed points and then weights the model based on that initial position. I am actually surprised that coping the construction points works at all.

I think what you are asking for is: Take this mesh and start FFD with this group of arbitrary positioned control points. That should be possible and probably not to difficult.

But you would have to be careful. When you start a FFD it creates a table of weights for each construction point with 3 floats per vertex in the group to deform.
So if you have a mesh with 1000 verts and a 3x3 set of control points:
3x3x3=27 control points
2710003=81,000 floats

When you update it loops through each vertex and calculates how much to move it based on the weighted changes in the control point. So the more control points you have the slower it is.

In your picture it looks like you have 9x9x3 control points. That comes to 729,000 floats. It would be really really slow to deform.

Chris

juju · 11 Feb 2008, 22:08

I'd take "really really slow" over "no such function at all" any day.

Pixero · 11 Feb 2008, 22:21

I think its pretty clear what he means.
He much rather have a slow tool that does what he wants then not having that tool at all.

CPhillips · 11 Feb 2008, 22:45

@unknownuser said:

The example in the pic takes a few seconds to deform.

So do the points placed actually have different coordinates assigned to them? That's what I was asking in my previous post. I find it kinda hard to explain, I hope you know what I mean.

Yes they do. Each construction point has its original position stored in an attribute. Thats how I know how far the point moved. But that is only 1/2 the equation. You also need the pre-calculated weight. That is what is missing when you copy a control point. Its that quick because it doesn't have all the calculated weights to process.

If you want to try it with more control points do this:
-Select the group you want to deform.
-Open the ruby console and enter the following line:

startFFD(Sketchup.active_model.selection[0],7.0)

That will create a 8x8 control grid (8-1=7.0). Delete the points you don't intend to move to make it a bit faster.

ffd9x9.skp

CPhillips · 12 Feb 2008, 07:22

I made a new version that has a NxN FFD. That means you can select how many control points there are in each direction. It isn't as slow as I expected.

Here is a 12x12x2. I am playing back my undo buffer so it looks faster than it is.

Ill post it once I fix a bug or two.

Just when I thought I was out they pull me back in.
Chris

juju · 12 Feb 2008, 10:17

@unknownuser said:

@unknownuser said:

I'd take "really really slow" over "no such function at all" any day.

Juju, I find your replies are often cryptic. What's the bottom line?

@pixero said:

I think its pretty clear what he means.
He much rather have a slow tool that does what he wants then not having that tool at all.

Exactly! Thanks for clearing that up for me.

Fletch · 13 Feb 2008, 06:29

Can't wait to play with this, Chris, thank you from the bottom of my heart... been wishing for this for a loooong time.

CPhillips · 13 Feb 2008, 06:53

This version supports NxN deforms. But more importantly it updates the mesh automatically when you move control points. That makes it a lot more usable.

Also added status text progress when it is weighing and deforming the mesh.

Making a curling wave with a 13x13x2 FFD.

See the PluginStore for the download...

remus · 13 Feb 2008, 07:00

cheers for the update chris, this will make an already powerful tool all the better.