Hi Jeremy,
Your number of iterations was too low. SurfSmooth -help hints at the issue and you can get more details on the topic from the references mentioned at the end of the help message.
Try that for amusement:
suma -spec mcgurk002_lh_mesh160_std.spec -niml &
then switch to the inflated view and turn on the recorder with 'R'
Now run the following (using the data you sent me):
foreach x (002)
foreach H (lh)
foreach LC (av.pa)
foreach Niter (100 500)
set icol = `expr 6 + 4`
rm -f "$H".160.sm8.mcgurk.deconvolution+reg."$LC"."$x"."$Niter"c"$icol".1D
SurfSmooth -spec mcgurk"$x"_"$H"_mesh160_std.spec \
-surf_A mcgurk"$x"_"$H"_mesh160_stdSmoothwm.asc \
-add_index -met LB_FEM \
-input "$H".160.mcgurk.deconvolution+reg."$LC"."$x".1D\["$icol"] \
-Niter $Niter -fwhm 8 -add_index \
-output "$H".160.sm8.mcgurk.deconvolution+reg."$LC"."$x"."$Niter"c"$icol".1D \
-talk_suma
ScaleToMap -input "$H".160.sm8.mcgurk.deconvolution+reg."$LC"."$x"."$Niter"c"$icol".1D -1 1 \
-cmap afni_n12 -anr 0 \
> "$H".160.sm8.mcgurk.deconvolution+reg."$LC"."$x"."$Niter"c"$icol".1D.col
end
end
end
end
You can watch the algorithm go into the oscillation mode with the 100 iterations and the steady mode with 500 iterations. The script will also generate color files like the ones you used for later display.
500 iterations might be overkill, try something like 200, it might work well for you.
Also, why do you generate meshes with such a high density ? I found an edge subdivision of 140 to be enough...
I will generate some pictures of all this some day.
cheers,
-z