archive/blender-archive
Archived
1
1
Fork 0
Code Pull Requests Packages Activity

Fix #33497: seting object scale to 0 on one axis made moving verts in edit mode

Browse Source 
impossible. In this case matrix inversion failed and didn't give a useful result.
Now it falls back to a pseudoinverse in that case, so that moving along the
other axes still works.

There may be other places that can benefit from this, but this is a place where
it has no significant performance impact, doing this in general for e.g. bone
matrices could be quite slow.
This commit is contained in:
Brecht Van Lommel
2012-12-14 21:41:22 +00:00
parent f1d487d1c4
commit 071a43f754
3 changed files with 31 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
source/blender/blenlib/intern/math_matrix.c
View File

@@ -504,8 +504,7 @@ void sub_m4_m4m4(float m1[4][4], float m2[4][4], float m3[4][4])
m1[i][j] = m2[i][j] - m3[i][j];
}
/* why not make this a standard part of the API? */
static float determinant_m3_local(float m[3][3])
float determinant_m3_array(float m[3][3])
{
return (m[0][0] * (m[1][1] * m[2][2] - m[1][2] * m[2][1]) -
m[1][0] * (m[0][1] * m[2][2] - m[0][2] * m[2][1]) +
@@ -534,7 +533,7 @@ int invert_m3_m3_ex(float m1[3][3], float m2[3][3], const float epsilon)
adjoint_m3_m3(m1, m2);
/* then determinant old matrix! */
det = determinant_m3_local(m2);
det = determinant_m3_array(m2);
success = (fabsf(det) > epsilon);
@@ -569,7 +568,7 @@ int invert_m3_m3(float m1[3][3], float m2[3][3])
adjoint_m3_m3(m1, m2);
/* then determinant old matrix! */
det = determinant_m3_local(m2);
det = determinant_m3_array(m2);
success = (det != 0.0f);
@@ -1903,3 +1902,16 @@ void pseudoinverse_m4_m4(float Ainv[4][4], float A[4][4], float epsilon)
mul_serie_m4(Ainv, U, Wm, V, NULL, NULL, NULL, NULL, NULL);
}
void pseudoinverse_m3_m3(float Ainv[3][3], float A[3][3], float epsilon)
{
/* try regular inverse when possible, otherwise fall back to slow svd */
if(!invert_m3_m3(Ainv, A)) {
float tmp[4][4], tmpinv[4][4];
copy_m4_m3(tmp, A);
pseudoinverse_m4_m4(tmpinv, tmp, epsilon);
copy_m3_m4(Ainv, tmpinv);
}
}