Folks -- probably Nick?
I'd like to be able to work with some FS sessions using uncompressed data, ie: not using mgz files. (Why? So that I can conveniently use some other tools for troubleshooting and inspection, tools that don't know how to uncompress mgz files).
Questions:
1. If I just try to gunzip mgz files, some of their names collide with the directories that mksubjdirs has set up. Historically, most of these dirs used to act as containers for per-slice COR files, so they don't seem useful anymore. Can they just be deleted without consequence?
2. I'm assuming that the uncompressed files should be given the extension ".mgh". Do the FS pipeline scripts and programs automatically adapt to uncompressed files?
Thanks,
Graham
Rather than using uncompress, you can just
mri_convert orig.mgz orig.mgh
Graham Wideman wrote:
Folks -- probably Nick?
I'd like to be able to work with some FS sessions using uncompressed data, ie: not using mgz files. (Why? So that I can conveniently use some other tools for troubleshooting and inspection, tools that don't know how to uncompress mgz files).
Questions:
- If I just try to gunzip mgz files, some of their names collide with
the directories that mksubjdirs has set up. Historically, most of these dirs used to act as containers for per-slice COR files, so they don't seem useful anymore. Can they just be deleted without consequence?
- I'm assuming that the uncompressed files should be given the
extension ".mgh". Do the FS pipeline scripts and programs automatically adapt to uncompressed files?
Thanks,
Graham
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
On Wed, 2006-09-06 at 13:43 -0700, Graham Wideman wrote:
Folks -- probably Nick?
I'd like to be able to work with some FS sessions using uncompressed data, ie: not using mgz files. (Why? So that I can conveniently use some other tools for troubleshooting and inspection, tools that don't know how to uncompress mgz files).
Questions:
- If I just try to gunzip mgz files, some of their names collide with the
directories that mksubjdirs has set up. Historically, most of these dirs used to act as containers for per-slice COR files, so they don't seem useful anymore. Can they just be deleted without consequence?
Yes, those directories (orig, T1, etc...) formally used to store COR files can be deleted without consequence (that is, assuming you have the .mgz or .mgh files).
- I'm assuming that the uncompressed files should be given the extension
".mgh". Do the FS pipeline scripts and programs automatically adapt to uncompressed files?
Yes, freesurfer recognizes the filename extension .mgh (and just doesnt do the gunzip on it internally).
Thanks,
Graham
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
1. You should keep the orig directory as that's where you store the the input data
2. I'm not sure that everything will run smoothly with mgh. I think a lot of stuff is hardcoded to work with mgz and/or COR
Nick Schmansky wrote:
On Wed, 2006-09-06 at 13:43 -0700, Graham Wideman wrote:
Folks -- probably Nick?
I'd like to be able to work with some FS sessions using uncompressed data, ie: not using mgz files. (Why? So that I can conveniently use some other tools for troubleshooting and inspection, tools that don't know how to uncompress mgz files).
Questions:
- If I just try to gunzip mgz files, some of their names collide with the
directories that mksubjdirs has set up. Historically, most of these dirs used to act as containers for per-slice COR files, so they don't seem useful anymore. Can they just be deleted without consequence?
Yes, those directories (orig, T1, etc...) formally used to store COR files can be deleted without consequence (that is, assuming you have the .mgz or .mgh files).
- I'm assuming that the uncompressed files should be given the extension
".mgh". Do the FS pipeline scripts and programs automatically adapt to uncompressed files?
Yes, freesurfer recognizes the filename extension .mgh (and just doesnt do the gunzip on it internally).
Thanks,
Graham
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
they will process the input based on the file name extensions, the problem is that things (like mris_make_surfaces) expect certain files (like brainmask.mgz) to be in certain places. This is the advantage of the simplicity of a script like recon-all: it hides all the complexity from you. But the price of course is flexibility - if you start changing names it won't know where to find things.
Bruce
On Wed, 6 Sep 2006, Graham Wideman wrote:
Doug and Nick:
Thanks for the quick replies. Notes:
Doug wrote:
mri_convert orig.mgz orig.mgh
Ah, good to know, and lets you rename in the same step, so avoids collision.
At 9/6/2006 02:14 PM, you wrote: 1. You should keep the orig directory as that's where you store the the input data
Yep, I knew about that one :-)
Nick said:
Yes, freesurfer recognizes the filename extension .mgh (and just doesnt do the gunzip on it internally).
But Doug said: 2. I'm not sure that everything will run smoothly with mgh. I think a lot of stuff is hardcoded to work with mgz and/or COR
Hmmm, so not sure what to make of that. Ideally programs would process the input based on the filename extension, and then produce same-format output. But that may not have been a priority.
Graham
freesurfer@nmr.mgh.harvard.edu