Problem with VASP and non-orthorhombic coordinates

[Aaron Deskins]

I am using the latest version of VESTA for mac. I have a simple VASP POSCAR file. It is non-orthorhombic. VASP calculates different cartesian coordinates than VESTA. I have manually confirmed that VASP is correct. Is this a bug in VESTA? Can it not handle non-orthorhombic VASP files?

Here's the VASP POSCAR file.
Ni O
   1.0000000000000000
     8.2802004322771552   -0.0083933393913823   -0.1072273514705764
    -0.0089139135642778    8.2824763273130184    0.2653513405601896
    -0.4556620900921381    1.1416993011562100   34.9045183394800134
   Ni
    1
Direct
  0.9906475751155099  0.0264665199582979  0.1764044882748692

VASP reports cartesian coordinates of 8.12214372  0.41229436  6.05811211.

VESTA reports cartesian coordinates of 8.042596    0.617469    6.146163.

The VASP coordinates are correct (verified by calculator). 

[Quang Nguyen]

I can confirm this.

Sometimes I get wrong value, sometimes get correct value of cartesian coordinates, depending on the lattice vector a1, a2, a3.

For instance:

---POSCAR_Test1---

Test1
1
 4 0 0
 0 4 0
 0 1 4
 Ni
  1
Direct
  0.5 0.5 0.5

===> VESTA: 2.00000   2.50000   2.00000
===> By hand: 2.00000   2.50000   2.00000

---POSCAR_Test2---
Test2
1
 4 0 1
 0 4 0
 0 0 4
 Ni
  1
Direct
  0.5 0.5 0.5

===> VESTA: 2.54662   2.00000   1.94028
===> By hand: 2.00000   2.00000   2.50000

Dr. Momma, please let us know if you have any comment on this?

Thank you.

[Koichi Momma]

In the current implementation of VESTA, unit cell parameters are calculated from the given 3 x 3 matrix, and then the structure is oriented such that the a-axis is parallel to the x-axis of the Cartesian coordinate system, and b-axis is parallel to the x-y plane.
For some file format such as *.xsf for XCrysDen, orientation of the structure will be kept unchanged.

In the latest development version, I also did the same thing for VASP format. So please try it.
http://jp-minerals.org/vesta/archives/testing/

Koichi

--

---
You received this message because you are subscribed to the Google Groups "VESTA users' list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vesta-discus...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/vesta-discuss/d4d7ba33-fb04-404f-b319-43679fa16081n%40googlegroups.com.

[Quang Nguyen]

Thank you very much for the explanation, Dr. Momma.

I have checked the development version (3.9.0a Oct 14) and it's ok now.

Best regards,

Quang Nguyen

[Aaron Deskins]

I'm glad to hear this new version works!