two spine deformers

[wesserbro]

Hi all! Need an advice on how to properly envelop deformation by two spine curves. I want the weight of a closest vertex to be 1, and horizontally the further one to be 0. Right now its defined solely by radius and doesnt count shape's wide and narrow parts. I can manually make variable radius by drawing slope on a radius panel but its really PITA even for such a simple figure.

want R2<R1 (not manually)

[Mathaeus]

Yeah it's radius in 3d space and that's it. While it allows to use Weight Editor, it's possible by occasion, better to do not try (from what I heard). There's ICE and modulating by weight map for such variances. While in ICE, you'll have to use a pair of reference - deformed curve.
If goal is just to snap a loop of vertices to curve, as plain modeling task, you can Shrink Wrap vertices to curve - if some closest location functionality in Shrink Wrap is used, like closest surface or closest vertex.

[wesserbro]

Hi, Mathaeus ) The actual task (maybe not very rational and quite enthusiastic one) is to control (for modeling purposes) human torso (or other parts) proportions with as few controls as possible. horizontal profiles (cross sections, transverse one) are defined and i dont want them to deform at all(only change proportions), so deformation should be equall to something like this: torso is sliced horizontally by infinite number of slices(cross sections), and each slice scales in horizontal plane to fit border curves.
Spine deform is giving me something suitable, and the only problem is the radius.
You can suggest cage deformer, but with control points on borders only (pic.2) its not good at all (distorting shape too much), by adding more points in between (pic.3) its still worse than spine deform and obviously i dont want to care about those additional vertices

Actually, yesterday i came up with an idea how to combine weights from two curves so that it would give me what i want, but i dont know how to access and change those clusters ))