Howdy-

Rick asked about this message:

+++ Arranging 3 images (each 416x560) into a 3x1 matrix.
You can view image ../init_qc_00_overlap_usrc_obase.jpg with:
aiv ../init_qc_00_overlap_usrc_obase.jpg
\n+* Removing temporary workdir 'o.aw_sub/__workdir_olap_rTdGepEA8Og*'\n

... it actually just comes from the initial QC image, showing the starting alignment. It is from the script @djunct_overlap_check, which is calling 2dcat here to glue a few images together to make "init_qc_00_overlap_usrc_obase.jpg". I don't expect this is a problem.

The @SSwarper script gets to Step #4 before crashing-- that is a 3dQwarper step, which might indeed involve a lot of memory. I actually am trying to process the same dset on my Mac with OMP_NUM_THREADS set explicitly to 3, just to see how that goes (I expect it should be fine, just slower than when using the default 8). To do this, in a tcsh script, you could have:

#!/bin/tcsh

setenv OMP_NUM_THREADS 3

@SSwarper ... [rest of command]

If you had a bash script, I think you would do this with:

#!/bin/bash

export OMP_NUM_THREADS=3

@SSwarper ... [rest of command]

--pt