Output IPF color map does not match given data

39 views
Skip to first unread message

Lila Hutchison

unread,
Apr 13, 2023, 10:50:49 AM4/13/23
to dream3d-users
I was given the IPF Color Map on the left from my professor's EBSD data. The image on the right is the one that I created by outputting an .xdmf file from Dream3D and opening it in Paraview. Would someone be able to explain to me why my output does not match the one given to me? I've attached a screenshot of my pipeline as well. I was given an .ang file that I converted from HEX to SQR, then to .h5ebsd, and then inputted into this pipeline. 
Screenshot 2023-04-13 104749.png
Screenshot 2023-04-13 104357.png

Michael Jackson

unread,
Apr 14, 2023, 9:36:32 AM4/14/23
to Lila Hutchison, dream3d-users
Dear Lila, 
    In ParaView there is a check box to “Map Scalars”, that is probably ON and it needs to be OFF and you are possibly NOT coloring by “IPF Colors”? Attached is a screen shot of ParaView correctly setup to display IPF Colors.

As a check you can also using the “ITK::Image Writer” filter to save the IPF Colors to a standard image file. Screen shot also attached of that filter.



Mike Jackson
BlueQuartz Software


On Apr 13, 2023 at 10:50:48, Lila Hutchison <lil...@vt.edu> wrote:
I was given the IPF Color Map on the left from my professor's EBSD data. The image on the right is the one that I created by outputting an .xdmf file from Dream3D and opening it in Paraview. Would someone be able to explain to me why my output does not match the one given to me? I've attached a screenshot of my pipeline as well. I was given an .ang file that I converted from HEX to SQR, then to .h5ebsd, and then inputted into this pipeline. 

--
You received this message because you are subscribed to the Google Groups "dream3d-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dream3d-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dream3d-users/787bf4e0-45b4-4706-8536-d4ba34980ecbn%40googlegroups.com.
Screen Shot 2023-04-14 at 09.31.42.png
Screen Shot 2023-04-14 at 09.34.21.png
Screen Shot 2023-04-14 at 09.35.10.png

Lila Hutchison

unread,
Apr 14, 2023, 4:30:56 PM4/14/23
to Michael Jackson, dream3d-users
Hi Mike, 

Thank you for your help! The map scalars checkbox fixed the issues I was encountering. 

- Lila
--
Virginia Tech Class of 2024
B.S., Aerospace Engineering 
Green Engineering Minor

Society of Women in Aviation and Space Exploration (SWASE) | Vice President
Be the Match at VT | President

Lila Hutchison

unread,
May 29, 2023, 2:14:53 PM5/29/23
to dream3d-users
Follow up question...

Would anyone be able to explain why the Euler Angle map is so noisy/grainy? The screenshot is the Euler Angle map for the same example as previous questions. 

Screenshot 2023-05-29 141312.png

Joe Wendorf

unread,
May 29, 2023, 5:46:19 PM5/29/23
to Lila Hutchison, dream3d-users
Hello Lila,

This is normal, and it is one of many reasons why we tend not to analyze Euler angles directly. To put it simply, there is more than one way to describe a particular crystal orientation with Euler angles, so two sets of Euler angles which have very different values may in fact be similar or even identical orientations. If this is a problem, and you need a single orientation to be described with a single set of Euler angles, try looking into any Dream3D filters which mention a "fundamental zone." However, I would recommend first looking into alternate ways to perform your analysis before resorting to this, as there are many other reasons why Euler angles are not ideal for analysis.

Regards,

Joe

Michael Jackson

unread,
May 30, 2023, 1:22:58 PM5/30/23
to Lila Hutchison, dream3d-users
Dear Lila, 
   You would want to color by “IPF Colors” which DREAM.3D has a filter to generate. The IPF Color scheme (and disorientation color schemes) were invented to help humans “see” the crystal orientation data much more easily than trying to colorize a raw Euler Angle. Can you color by “IPF Colors” inside of ParaView and send that image instead? I think you will find the results much more workable.

--
Mike Jackson                    mike.j...@bluequartz.net
BlueQuartz Software         www.bluequartz.net
President/Owner               Dayton, Ohio
Principal DREAM.3D Developer


Reply all
Reply to author
Forward
0 new messages