cannot create new phases

[Voslab]

I installed EBSD-Image from the repository on a Ubuntu 12.04 64 bit computer.
When trying to open Create new phase I get the following message:
"Error!!!
Source: rmlshared.thread.VolatilePlugin
Exception: java.lang.UnsupportedOperationException"

I tried to load some .cif files, created by GDIS or from the Crystallography Open Database but none was succesful.

[Voslab (adrian-voslab)]

Same behaviour (same error message described as appearing in Ubuntu) on Windows 7 on the same computer (dual boot)

[Voslab (adrian-voslab)]

Just to let you know that the above erros occur in the latest version 0.2.5.
I've tested version 0.2.2 on both Win7 and Ubuntu 12.04 and it seems that everything works like a charm!

[Philippe Pinard (philippe.pinard)]

Thank you for the detailed description of the bug. I was able to found the problem and fix it. I will create a new version soon. It will be 0.2.6.

[Philippe Pinard (philippe.pinard)]

The new version is up. Concerning the CIF files, could you send me one or two examples that does not work? I'll try to find the problem with the parser.

[Voslab (adrian-voslab)]

Thanks for your swift response! Attached is a .cif file - converted by GDIS (http://gdis.sourceforge.net/) from GULP format. I also tried a file from the Crystallography Open Database: COD ID: 9012188.

[Voslab (adrian-voslab)]

I've tried version 0.2.2 and almost everything went fine: simulated patterns are OK, importing from HKL no problem, running an experiment produces clear peaks on the Hough maps but no indexing, no phase identified. Maybe I'm missing something. I'll try again, maybe v. 0.2.6. Thanks

[Philippe Pinard (philippe.pinard)]

The indexing is not part of the main distribution as it is too experimental at the moment.

[Voslab (adrian-voslab)]

Just tried v.0.2.6 and ...it works! Indexing using the Krieger-Lassen algorithm. I've got a phase map very similar to that obtained using Channel5 for a material with 3 phases.
For HKL imports you need to adjust some values in the .cpr file. If anybody was interested I could post them somewhere - not here, this is about bugs.

[Philippe Pinard (philippe.pinard)]

Glad to hear that you got a decent indexing. I'm actually very surprised!

For the modifications of the .cpr file, are these major modifications? If so, I can probably make them internally when it is loaded inside EBSD-Image. If not, I can add your recommendations to the online documentation (http://ebsd-image.org/documentation/beginners/import/hkl.html). Let me know.

If you want, you can send them to me at <email address hidden>.

[Voslab (adrian-voslab)]

Well, it wasn't great but not bad either. Sorry, I was too excited in my previous comment. I have 3 phases: a cubic phase, a tetragonal and an orthorhombic phase. The orthorhombic is only about 3 percent so when indexing with all 3 the orthorhombic phase takes the entire map. If I don't include it the cubic and tetragonal are indexed correctly, I think. I attach one solution for the tetragonal so you can have a look. I selected this pattern because is rather poor but the solution seems to be OK even with this.
Regarding the HKL import. In the .cpr file the pattern center is calculated starting from the bottom-left corner while in EBSD-Image is calculated starting from the top-left corner. So we need to change the PCX and PCY values in the cpr file. I've noticed this by comparing the PC in Channel5 and EBSD-Image and then I checked Channel5 manual (Release date July 2001) page 6.32.
Thanks for your help

[Voslab (adrian-voslab)]

It's probably good to see the pattern without solution overlay. Attached