JEOL Map import

[Heather Lowers]

Any clever people out there written a script to take maps collected with, say an 8900, read the condition files and ascii file to create a file that could be opened in Probe Image or Calc Image?  Ideally would like to quantify old maps.

[Probeman]

Any clever people out there written a script to take maps collected with, say an 8900, read the condition files and ascii file to create a file that could be opened in Probe Image or Calc Image?  Ideally would like to quantify old maps.

Hi Heather,
Several years ago when Probe Image was still "a little rough around the edges" for certain types of mapping acquisition, Philippe Pinard at Aachen wrote a python script (I think) to convert x-ray maps from the JEOL 8530 software into GRD files that CalcImage (and Probe for EPMA) could quantify.  But I don't know of any scripts specific to the JEOL 8900, unless JEOL is still using the same file format for x-ray maps- which they might very well be.  I would start with asking him, though he works for Oxford Instruments in England now.

Since you mentioned quantifying old x-ray maps, I have to wonder if the x-ray intensities (maps, but also standard intensities), stable enough over time to quantify them using more recent standard intensities?  I think trace element maps would be OK (assuming you have off-peak maps), but major elements could show large divergences if your instrument intensities have drifted over time by very much...

The other possibility would be to figure out a way to import your old JEOL maps into Surfer for conversion to GRD files.  Surfer supports many data formats including some raster formats. Here is a screen shot of a few of the formats it can convert. Maybe one of the image formats would work?


john

[Philippe Pinard]

Heather, John,

Can you send me one .map file from your 8900? I need to check if it is the same as the 8530. Also is there a .cnd file for each map?

philippe DOT pinard AT gmail DOT com
Philippe

[Ben Buse]

I wasn't sure where to post this. But...

John now includes a conversion from JEOL or Cameca maps to PrbImg in Probe Image.

(1) File menu - Convert JEOL or Cameca to PrbImg.
(2) Convert JEOL 8530 to PrbImg (i.e. my instrument)

I post this because - I was using it the other day but couldn't figure out what I was doing wrong - and thanks to John's help - I thought I'd post it in case anyone else has the problem

The key is to follow the instruction



When it says click on 'map.cnd' file it means the 'map.cnd' file, NOT any map file with a .cnd filename

Ben

[John Donovan]

Hi Ben,
Glad to hear it's working for you now.   

For those who are wondering why Ben doesn't just utilize Probe Image for all map acquisitions, it's because Probe Image doesn't yet acquire polygon maps nor does it yet support the image (stage) drift feature. So until we implement these features directly in Probe Image, we've provided a convertor utility accessed from Probe Image, to allow one to acquire x-ray maps using the JEOL or Cameca OEM software, and then convert the maps to the Probe Image PrbImg file format. These PrbImg files can then be utilized by CalcImage to create fully quantitative x-ray maps automatically corrected for beam drift, standard drift, background, matrix and spectral interferences, as shown here for example:

http://probesoftware.com/smf/index.php?topic=73.msg5130#msg5130

Ben: yes, the map.cnd file contains the mapping information for all the maps acquired in that session., so that's the map file that needs to be clicked on for the JEOL 8230/8530 map import and conversion to PrbImg file format.

For the JEOL 8900/8200/8500 one needs to click on the 0.cnd file, as that is the file that contains the mapping information for all the acquired maps for these instruments.

For Cameca instrument map files one needs to click on any of the ASCII .TXT files that has been exported from PeakSight with a caveat: PeakSight does not provide quite enough information in its export files to uniquely determine the exact identity of each file to be converted. So depending on the naming convention utilized by the user, one may or may not need to rename the files to be sure the x-ray and VS signals are assigned properly. It's hard to explain but in looking at the Cameca .TXT file, it seems that Cameca never intended for this .TXT format to be automatically read in by a re-processing program.  So if anyone out there has figured out the actual header format of the Cameca binary mapping file (*.ImpDat),  please contact us and let us know what you know. Cameca has refused so far to share any information on their binary format, because well, they're Cameca.   

Also, for anyone else wanting to utilize the File | Convert JEOL or Cameca files to PrbImge format menu, be sure to download the latest ConvertToPrbImg.exe file and copy it to the Probe Image application folder and overwrite the existing file in the Convertor sub folder. 

We've fixed some minor bugs in it since the last release of Probe Image.  I've attached it below in case anyone needs it.

[Hwayoung Kim]

This is really cool cause we frequently acquire polygon maps using JEOL software.

I have just downloaded the latest ConvertToPrbImg.exe file and tested the conversion. When I select individual map files (e.g. data001.cnd, data002.cnd), each maps are correctly converted to prbimg file. But when selecting 'map.cnd' file which locates in a same folder with the map files, the software looks for a data010.map file even though there are only 9 map data. (I've attached a screenshot.) Do I miss something?


Hwayoung

[John Donovan]

This is really cool cause we frequently acquire polygon maps using JEOL software.

I have just downloaded the latest ConvertToPrbImg.exe file and tested the conversion. When I select individual map files (e.g. data001.cnd, data002.cnd), each maps are correctly converted to prbimg file. But when selecting 'map.cnd' file which locates in a same folder with the map files, the software looks for a data010.map file even though there are only 9 map data. (I've attached a screenshot.) Do I miss something?

Hi Hwayoung,
You should be selecting the map.cnd file as this file contains the information for all the acquired maps.  Can you send me the map.cnd file so I can see what is going on with the 10th map?   In fact please send me all the .cnd files from this acquisition and I will run through the conversion myself and check if there's something that needs to be tweaked in the code.
john

[JonF]

I was just about to post that I was seeing the same issue with PI looking for a 10th map file when I had 9 old JEOL maps to convert, but I just updated the ConvertToPrbImg.exe linked above and it sorted my issue - nice! Thanks, John!

[Hwayoung Kim]

John had fixed the issue on importing JEOL 8230/8530 map, and now it works perfectly for our data. Thanks, John!

We just need to select 'map.cnd' file in the same folder where all elements' .map and .cnd files locate, and ConvertToPrbImg.exe converts all JEOL maps to .prbimg properly.

Currently the only difference between .prbimg files acquired by PI and converted from JEOL map is that the latter have the same sequential number in their filenames for all elements.

In this example, we have acquired 10 elements x-ray maps through separate passes (Na, Si, K, Fe, Cr first and Mg, Al, Ca, Mn, Ti second) using JEOL software and convert them to .prbimg files (left in the following file lists). They all have same sequential number “00001” in the filename. If same measurements were done using ProbeImage, sequential numbers will be different for the elements of different passes like the example on the right.



This makes a slight change in the CalcImage map quantification procedures of Karsten's posts. (https://probesoftware.com/smf/index.php?topic=146.0) At the step when selecting .prbimg files to import (when Open File dialog opens), we only have to select any elemental map ONCE (not twice to select both passes) and then to click "Cancel". Then CalcImage loads entire maps and convert them to .grd files at once since they have same sequential number. The other steps are exactly identical to the original procedures.

Below is an example of our quant map data acquired by JEOL software, converted to .prbimg by ProbeImage, and quantified by CalcImage.

[Probeman]

Hi Hwayoung,
Very nice, thanks for posting this subtle but important point.  Nice quant maps also.

On the color scale label format, I have found that specifying the "Compact" label format option and 3 digits works really well for almost any data range of the Z data:



john

[Mike Spilde]

I'm trying to run the ConvertToPrbImg app. After a false start, I have downloaded the latest version (Aug 2018) and followed John's instructions but still can't get it to run. Now I'm getting an error: "Invalid invocation of ConverToPrbImg executable. Contact Probe Software technical support"
Can anybody offer some "technical support"?
Thanks

[John Donovan]

Hi Mike,
The technical support contact for Probe Software is support@probesoftware.com, but since you posted your questions here on the user forum I'll respond here.

Are you trying to run ConvertToPrbImg.exe by double-clicking on it?  That will not work.

The ConvertToPrbImg.exe file needs to be copied to the C:\Program Files (x86)\Probe Software\Probe Image\Converter folder. And then you need to launch it from the Probe Image File | Convert JEOL or Cameca Files to Probe Image Format menu.

If you run the latest ProbeImage.msi installer, the app will get installed to the proper folder automatically. But I've attached the latest version below if you want it.   Let me know if you still have issues running it.
john