Hi, thanks very much for your help guys. I've processed these amounts of TRs and runs before and its usually much quicker than this.

I noticed in the output that epi is being resampled to match anat. Given that the anat has a 1mm res (vs 2.5x2.5x4 for the original epi), I'm guessing that this might be slowing things down during 3ddeconvolve.

I've provided the output here -->

++ Deobliquing original dataset before obliquing
++ Combining oblique transformations
++ Output dataset ./__tt_CRD.001_anat_ns_ob+orig.BRIK
#++ Creating representative epi sub-brick
#Script is running (command trimmed):
3dbucket -prefix __tt_pb01.CRD_001.r08.tshift_ts ./pb01.CRD_001.r08.tshift+orig'[179]'
++ 3dbucket: AFNI version=AFNI_2011_12_21_1014 (Apr 4 2012) [64-bit]
#++ resampling epi to match anat data
#Script is running (command trimmed):
3dresample -master ./__tt_CRD.001_anat_ns_ob+orig -prefix __tt_pb01.CRD_001.r08.tshift_ts_rs -inset ./__tt_pb01.CRD_001.r08.tshift_ts+orig'' -rmode Cu


I can't work out where I've stated that this should occur. Could you let me know if there is an option to change this at the afni_proc.py stage? This is what I have for the align stage of my proc script -->


# ================================= align ==================================
# a2e: align anatomy to EPI registration base
# (new anat will be aligned and stripped, CRD.001_anat_al_keep+orig)
align_epi_anat.py -anat2epi -anat CRD.001_anat+orig \
-suffix _al_keep \
-epi pb01.$subj.r08.tshift+orig -epi_base 179 \
-volreg off -tshift off


Thanks!

David