Every time I think I have the curve modifier mastered, it throws me a another curve ball (no pun intended).

Why in this blend file can I not get the lights to follow the spiral curve? When the modifier is activated, the lights are stretched and distorted along the curve.

I created the curve by converting it from a mesh (string of verts). Why does it not display like a normal Bezier? Is that related to my problem?

I am following all the "rules" that I am aware of, as detailed here: Understanding the Curve Modifier. The curve seems to run in the correct direction. The origin point is at the first point of the curve. The curve modifier on the lights is set to -Y which is the local orientation of the curve's direction.

What am I missing?


1 Answer 1


General Tips

A few general guidelines regarding Bezier Curve Objects and the Curve modifier.

  1. For more predictable results set curve origin to one of its ends, preferably its starting point.
  2. If necessary under Edit Mode invert spline direction so the start point is at the same end as the origin from Segments > Switch Direction.
  3. Place both the curve and the deformed object at the same location in the scene so that both centers match in world coordinates.
  4. Make sure both objects have matching rotations on all axes. Clear rotations beforehand, correctly orient your mesh, and then apply all rotations in Object Mode. Afterwards adjust as necessary, rotating both curve and mesh in tandem.
  5. Apply scale on both objects. Having uneven or unmatching scale factor will lead to issues.
  6. Make sure curve bevel radius is uniform and at a value of $1$, anything else will introduce distortions or length mismatches.
  7. If you apply scale, it will multiply bevel radius by the old scale and you will need to set the radius back to $1$ in Edit Mode, available from the properties sidebar (Nkey) or by searching for Set Curve Radius. Applying scale doesn't clear the bevel radius.
  8. Make sure your mesh extends roughly in the same direction the curve object, growing along its side, otherwise it may produce no visible results.
  9. If you get deformations along the wrong direction adjust the deformation axis of the Curve modifier to match the direction of your curve and mesh object.

None of these individual steps are strictly mandatory, but they certainly help get more predictable results. Unless you know what you are doing they may help get started.

After a workable starting point you can adjust as necessary, breaking any of the above points you deem necessary to get the desired results.

In this particular case

First you should sanitize both your curve and your lights object. Correctly place your lamps exactly at the origin of the curve, clear its rotation, and then rotate it again so it faces the desired direction. After that apply the rotation.

Clear object transforms

There are a few types of different curve objects in Blender, Bezier Curves, NURBS Curves, or Poly Type curves. Since you converted from a mesh object it is by default a Poly type curve, which has no curvature and hence no handles. It can only contain straight "poly line-like" segments.

This has absolutely no influence in your Curve modifier bad results, and can easily be fixed by changing curve types. Under curve Edit Mode use the button on the toolbar that reads Set Spline Type > Bezier.

Change curve type

After that make sure you then change all handle types to Automatic or something other than vector to get a smooth deformation.

Then you can finally activate the Curve modifier. Switch its direction to X axis.

Anyway, the root cause of the weird results you are getting in this particular case are caused by having your curve vertex with a very small Radius factor, causing the mesh to be scaled along the curve. You can fix this by entering Edit Mode on the curve, in the Properties Shelf > Transform > Radius set it to 1 on all vertex. You can quickly override all selected vertices with the operator Set Curve Radius.

Curve radius

  • 1
    $\begingroup$ Perfect, thank you. Very comprehensive and well explained. Also leads me to believe that the information in that article I linked to in my original question is totally inaccurate. $\endgroup$
    – Lewis
    Commented Oct 22, 2016 at 6:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .