SoftEdit does it’s magic based on control points.
You need to prepare the surface to have the control points you need.
If you want to edit a surface from a location that does not have a point, add one (InsertControlPoint) or Rebuild to add the control points you need for what you want to do.
Ok… so why is the distance measured in world units rather than control point numbers then?
Do I dare to say that this is yet another example inconsistent behavior?
I mean, if my control points curve a lot (say even more than in the example above), is the world unit distance along the cplane x, y, z or does it actually measure the length of the iso curve? Not even the official Rhino documentation makes this clear.
The shape of a NURBS curve or surface is determined on the number and location of Knots and their derived control points, a weight associated to the control point, and the degree of the span.
SoftEdit is directly moving selected control points and then moving the surrounding points by a smaller distance to spread out the change. The new location of the points changes the shape of the curve/surface.
I think you missed my point. My suggestion is that the tool should not directly move an initial selected control point, but rather spread out the change from your initial surface click, that’s all. It’s only the initial snapping I object to, because it reduces the the potential of this tool and forces the user to make pre-emptive changes to the surface they might not desire.
I think you fundamentally missed My point.
It works by moving control points. It finds the closest one to where you pick.
Knowing that, you can prepare the specific surface by inserting knots, or some other method of adding control if needed.
in that case the tool needs to be revised. It makes no sense that the surface needs to be adjusted to the tool. But it makes complete sense that you would want to soft-move from the middle between 2 control points.
No, for SubD we have SoftTransform, which btw is also is a UX disaster. What we would actually need is falloffs that can be manipulated in 3D space. That stay visible and active until you shut them off. That are tool and geometry independent.
Sorry @John_Brock may I ask: what makes you say this?
If you don’t see the application of what is asked for here, please have a look at the following video (starting from 1:43)
P.S.: I know that the video is NOT showing a fallow that is independent of cv position, but it shows the advantage of defining the falloff distance in terms of cv numbers.)
This thread is only about the fact that despite using world units, there is an initial “snap to control point”, which to me seems to run counter to using world units in the existing SoftEditSrf command.
I’m not sure if I’m hitting a language barrier with Mr. John Brock again (English is not my native tongue), but perhaps @Pascal or @wim can chime in and see if they understand what I’m trying to say?
3DS Max, Z-Brush and other modeling programs have tools to soft-edit surfaces at the desired portion of a surface for a reason. They allow a surface to be edited exactly there the modeler wants.
It’s crucial for this command in Rhino to be able to work in the same convenient and predictable way, because the current snapping to a control point far away from the desired area is misleading and leads into unwanted results. Having to rebuild the structure of a surface to overcome that limitation is also highly unwanted, especially for people who like to keep their surfaces clean and easy to edit.
Snapping to the nearest control point also makes it impossible to soft-edit a single-span surface relatively symmetrical, because it always snaps away from the middle of the surface edge.
