About Scaling Rail

Next
 From:  Frenchy Pilou (PILOU)
2681.1 
Is that normal that gives not a result in a reasonnable time?
(in fact I don't know if it's possible :(

---
Pilou
Is beautiful that please without concept!
My Gallery
Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Mip (VINC)
2681.2 In reply to 2681.1 
Bonjour Pilou,

Shouldn't your scaling rail be an angled curve parallel to the curve rail instead of a straight line ?

 In fact, it works (9 seconds here), but what a surrealist shape %-)))
I also unchecked Constant Height.

-Michel

EDITED: 28 May 2019 by VINC

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
2681.3 In reply to 2681.1 
Hi Pilou - that's not really a proper setup for a scaling rail... Probably it is generating a very strange warped and chaotic result with a really large number of points in it.

A scaling rail for a 1-rail sweep is meant to follow more along the direction of the path, and also "cover" it completely - meaning any perpendicular you trace out from the path should intersect the scaling rail instead of shooting off into space. It looks like in your case there the scaling rail does not "cover" the entire path.

Here's some illustrations for how the scaling rail works - if you have a sweep with the scaling rail off to the side, like this:



Then when you apply the scaling rail, the sections will attempt to extend along the perpendicular of the path, until they hit the scaling rail like this:




If there is not any intersection available (which I think is the case in some areas in your example), then the original sweep is used for that profile instead, but that will tend to be a very sudden violent change in shape in the surface and will make for some bad results.

Also in your case it looks like the scaling rail is not in the same plane as the path curve, for the most predictable results I would recommend putting the scaling rail in the same plane as the path curve for a 1-rail sweep.


If the scaling rail is not set up in a good relation to the original path curve, you can get a really messy and chaotic result, usually when things take a long time it means that is what is happening - you've got some strange surface result that has a whole bunch of swoops and bulges in it which tends to take a long time to generate during the refinement process and also can take a long time to generate a display mesh for.

So anyway, I hope the illustration above will show you more how a scaling rail is intended to be shaped!

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
2681.4 
Of course I have no problem when my Rail scaling is // and have some points down and on the Z axe
And also because I have made that with a curve line who have also only 2 points over the path on the Z! -works fine
And want to know if it's the same with a straight line

I have just asked for know if it was pertinent or not

In fact I suppose that this rail scaling can be considered as a sort of third rail?

EDITED: 12 Jun 2009 by PILOU

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
2681.5 In reply to 2681.4 
Hi Pilou,

> In fact I suppose that this rail scaling can be
> considered as a sort of third rail?

Yes, kind of... It works though as a modifier to the sweep, the sweep will initially be formed with the regular rails you give (either one or two rails), and then if you set a scaling rail, it will modify all the generated profiles of the sweep to stretch out in the way that I showed in that previous illustration.

The "main rails" of the sweep control the placement of each cross-section more directly - the "scaling rail" just controls a stretching factor.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
2681.6 In reply to 2681.5 
My line of the first post was a perversion of the method pushed to the limit :)

Ps Not possibility to have the Rail Scaling only limited of the "upper" half side of the volume?
(don't want the bottom section follow the Rail Scaling)
The funny thing is that you can move after the points for adjust your form!

EDITED: 12 Jun 2009 by PILOU

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
2681.7 In reply to 2681.6 
Hi Pilou,

> Ps Not possibility to have the Rail Scaling only limited of
> the "upper" half side of the volume?
> (don't want the bottom section follow the Rail Scaling)

Not with a scaling rail - it works by applying a scale transformation on the profiles, which does not easily allow for that kind of a "partial" scale in a general way.

But if I understand you correctly, you can get that kind of a shape by using a 2-rail sweep instead of using a scaling rail.

That would be a profile and rails like this:



Then when you do the sweep, select both of those as sweep rails at the rail selection prompt, instead of picking one as a scaling rail later. With both of those as sweep rails, the profiles will slide along the 2 rails which gives this kind of a result:



- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
2681.8 In reply to 2681.7 
yes but as my first volume is a 2 rails sweep (so Horizontal asymmetric side) maybe I can make that if the Rail Scale don't make half section
(not so a big deal :)

EDITED: 13 Jun 2009 by PILOU

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
2681.9 In reply to 2681.8 
Hi Pilou, also when you start wanting to get a lot of control in many different areas like this, you may want to switch to the Network command instead. There you can add in a larger number of profiles.

But also it looks like assembling your object in halves works for your case too?

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  Frenchy Pilou (PILOU)
2681.10 In reply to 2681.9 
yes assembling works fine!
I was afraid that horizontal section shoud be different after the rail scaling but all works fine :)
---
Pilou
Is beautiful that please without concept!
My Gallery
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All