Scale tool inferring, not eyeballing


I’m trying to reproduce the geometry of a “duck shaped” wine decanter.

To do this, I’ve traced the contours of the glass body, and have found the centre axis of its shape. From this, I’ve Follow-Me’d a decagon, or 10-sided circle, along said centre axis, and used the Scaling Tool to reach the topmost contour line in the volume.
The problem is, I can only eyeball the scale of each plane, but I’d need to infer it in order to obtain manageable curves, i.e. products of formulaic construction, in order to build tenable surfaces from these.

Is there any way to go about this?


Eyeballing is not necessarily with the scale tool , you can use references, scaling factors and specific dimensions in units. The difficulty with the scale tool is choosing which grip handles you should grab and by selecting only one surface, you would get lesser options and thus better control.
The problem with rotated surfaces is that the scale tool doesn’t recognize it as a plane. You would have to group it first in the Red-Green plane and then rotate the group. The scale tool comes with the ‘Toggle Uniform’ button ( Shift on Mac, see the status bar for options) and the group you’re scaling will be scaled uniform in all the planes ( because you have made a group of a face, there is only one)


The problem I have is that I’ve got to start from the central axis of the construction, since the upper and lower-most beziers are completely different.
I’ve used follow-me, a bezier curve and perpendicular face from a plug-in (icon:) to obtain the angles and spacings of each decagon I need to rescale.
What I think I’ll have to do is rotate the entire model for each scaling, so that the scale tool “sees” the selected group as a plane.
Any thoughts?
Sorry for the probably blurry language; These concepts are hard to communicate with words…


It worked!
I can do it one decagon at a time, provided I leave a reference line to enable me to return the face to its original angle once I’ve rotated, exploded and regrouped it. For some reason, the apexes aren’t lined up with the destination bezier, but scaling too big, leaving a reference at the intersection with the surface, and scaling back down makes inferring possible.
I’d say this is solved…


Make the first decagon a group that is oriented on the ground plane. Then copy it and place it along your curve with the needed rotations. Then edit each one in turn and scale the geometry inside. This will eliminate the need for reference lines and all that rotation. After you are satisfied with the decagons, explode the groups and move on to skinning the shape.


Reviewing the posted GIF file, I can’t tell how you get a scaling of the entire polygon after using a non-corner handle to obtain the desired dimension. What am I missing?


Throw in the Shift Key.
If you look here you see, scaling in one direction, then with ctrl scaling from center, and finally with ctrl and shift uniform scaling from center.


For some strange reason, I thought the Shift key only worked from corner handles.
Go figure…
Thanks a million…


This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.