EDAX/TSL x pattern center reversed in simulated patterns

[Habakuk224]

Hello,

I am trying to reindex a material using DI. However, the DI result is approximately 5° off the original data indexed by Hough indexing in OIM Analysis. After testing of various processing steps, I have discovered, that the pattern center defined in DItutorial for EDAX systems is probably reversed in x direction. An example follows:

A 120x120 simulation was done by EMEBSD for Euler angles 90, 22, 0 using

xpc = 30
ypc = 0
L = 26100

for BCC material, i.e. the [100] axis should be perpendicular to the detector. The xpc=30 is a bit extreme, but the error is then easily seen (while for typical xpc close to 0, the effect is barely noticeable). According to the definition from the DItutorial, this corresponds to

x* = 75 %
y* = 50 %
z* = 81.86 %

However, when viewed in OIM Analysis (v8) the pattern center defined in such way is shifted from the actual pattern center position on the [100] axis:

Correct pattern center is shown for x* = 25 %

Y pattern center is behaving correctly and ypc corresponds to its y* counterpart.

[marcdegraef]

Hi,

this was documented for EMsoft 5.0 on https://github.com/EMsoft-org/EMsoft about half-way down the page:

*

      *For Dictionary Indexing (DI) users, we have changed our
      convention for EBSD pattern orientation to be the same as the
      vendor convention; i.e., EBSD patterns are displayed looking
      from the detector to the sample. The practical consequence is
      that for older DI runs the sign of the xpc (x-component of
      pattern center) will need to be reversed when used with the
      Release 5.0 indexing programs.*

Regards, Marc.

On 6/3/22 12:11 PM, Habakuk224 wrote:

Hello,

I am trying to reindex a material using DI. However, the DI result is approximately 5° off the original data indexed by Hough indexing in OIM Analysis. After testing of various processing steps, I have discovered, that the pattern center defined in DItutorial https://github.com/EMsoft-org/EMsoft/wiki/DItutorial#51-setting-the-detector-geometry for EDAX systems is probably reversed in x direction. An example follows:

A 120x120 simulation was done by /EMEBSD/ for Euler angles 90, 22, 0 using

|xpc = 30 ypc = 0 L = 26100 |

for BCC material, i.e. the [100] axis should be perpendicular to the detector. The xpc=30 is a bit extreme, but the error is then easily seen (while for typical xpc close to 0, the effect is barely noticeable). x30_y0_E90-22-0 https://user-images.githubusercontent.com/72706923/171902105-758ec3b8-2d32-4818-af40-aaea7d5ec8e3.png According to the definition from the DItutorial, this corresponds to

|x = 75 % y = 50 % z* = 81.86 % |

However, when viewed in OIM Analysis (v8) the pattern center defined in such way is shifted from the actual pattern center position on the [100] axis: nok https://user-images.githubusercontent.com/72706923/171902134-7267de64-0063-4c85-891d-929ac56964cf.png

Correct pattern center is shown for x* = 25 % ok https://user-images.githubusercontent.com/72706923/171902128-1dc4b9b1-8612-4181-9b0b-2fd36b820dfe.png

Y pattern center is behaving correctly and ypc corresponds to its y* counterpart.

— Reply to this email directly, view it on GitHub https://github.com/EMsoft-org/EMsoft/issues/161, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB26VWEXDFXYIOA5XSSNQH3VNIVDBANCNFSM5XZPIIXA. You are receiving this because you are subscribed to this thread.Message ID: @.***>

[Habakuk224]

My mistake, thank you for clarification.

Regards, Jiri Kozlik