#51 Align to Selection issue

open
None
5
2012-10-25
2001-12-10
Anonymous
No

Problem: Creation of unflat planes and mis-alignment
anomolies.

Hesitant to refer to this as a bug, as I may be trying
to do things that it is not supposed to do.

After using Align to Selection on a flat face, I
thought that move free should allow movement of the
verts belonging to that face, around on the same plane
as the face.

This appears to be the case in fact if moving any of
the original verts.

However, if any one, or all, of the edges are cut, and
the new resultant verts moved around, then the face no
longer stays flat, although still subjected to AtS.

Try this on any face of a default sphere. I used faces
in the northern hemisphere that were not on a major
axis.

Also, select a face on sphere and use extract region
normal on it.

Reselect the face on the sphere and apply Align to
Selection.

In ortho mode, surely the extracted face should
exactly cover the original face on the sphere? It does
not.

There is also a slight movement between using AtS on a
face, and using AtS on the edges surrounding the same
face.

Other anomalies using AtS in edge mode on edge loops
of 3D twisted pipework type models.

A file would be needed to describe same properly, so I
will wait until (if) asked as I may be barking up the
wrong tree anyway.

pp

Discussion

  • Bjorn Gustavsson

    Logged In: YES
    user_id=286649

    I forgot about ortho mode. It calculates the view as if the
    view was perspective. I'll fix that.

    BTW, I see this command as temporary command, until more
    powerful commands are available.

    I think that the edges behave correctly. The Ats command
    calculates the normal for each edge, then averages them. The
    normal for an edge is calculated as the average of the
    normals for the faces on either side. Therefore I would
    expect Ats on a face or the same edges to have slightly
    different results.

     
  • Nobody/Anonymous

    Logged In: NO

    Re ortho mode fix.
    I assume that after this fix, then extracting normal a
    face from a previously aligned (AtS) face will not change
    that face in any way from the original as it is brought
    towards the camera? Exact fit and orientation?

    AtS with face, even when used as intended, seems odd.

    Default cylinder -> select top face -> rotate X 45 deg. ,
    use AtS on top face.

    I would expect to be looking at the true shape of the top
    face (circle), not what is actually presented?

    If I understand edge AtS implementation properly, then
    using the face AtS version seems like a safer bet anyway
    for doing some of the ‘slide’ type ops. Especially if both
    faces forming an edge are very dissimilar in shape and
    flatness?

    When you say temporary command, I assume you mean with
    reference to what is being discussed here. Once proper 3D
    non axis aligned transformation commands are available then
    there will be no further need for the likes of myself to
    mess around?

    I trust AtS will remain after implementing the big guns?

    I find it very useful for things like ears, leave a
    reference face at whatever plane angle other work is being
    done, knowing that use of AtS on that ref. face will always
    re-align things in the future?

    Ready made simple datum, in my books?

    pp

    Interesting stuff for the future, no wonder the natives are
    getting restless?

     

Log in to post a comment.