> I'm interested in averaging a couple datasets in SUMA space.
> From the documentation it looks like I need to map all my
> subjects to Icosahedron space,
This is not quite right. You are warping the meshes to either sphere.reg (typical for FreeSurfer) or sphere (in special cases). So your surface meshes will be relating to the spherical template used by FreeSurfer to create sphere.reg
> extract a functional dset for
> each subject (3dVol2Surf), then I can use 1deval to average the
> subjects.
Correct.
>
> First, I'm wondering if running MapIcosahedron space on each
> subject yields a common surface.
For each subject, you get 'standard mesh' versions of the original surfaces. Those std meshes look the same as the original ones (in terms of geometry) but the mesh has been changed in a way that will be consistent across the various subjects. You would use std-meshes only from this point on.

> Should I be able to load
> subject 1's dset onto subject 2's Icosahedron surface (without
> losing alignment)?
Not quite. Subject 1's volume data should now be mapped to subject 1's std-mesh surfaces.
Subject 2's data go to subject 2's stdandard meshes also.

Note that the std-surfaces from each subject are not aligned together. But node N from subject 1's mesh corresponds to node N from subject 2's mesh. That is why you can node-based group analysis.

> Or do I need to further align each subject's surface?
> From the documentation, I'm assuming that the averaged dataset
> could be loaded onto any subject's Icosahedron surface (so I'd
> think individual datasets should be able to be loaded across
> subjects).
If by average dataset you mean an average dataset created on the standard mesh (what you refer to as Icosahedron surface) then yes you can display a surface-based dataset defined on the standard mesh on any subject's std meshes.
>
> So far I have run MapIcosahedron on all my subjects and
> extracted dsets using 3dVol2Surf. When I load subject 1's dset
> in onto subject 1's Icosahedron surface, the result looks the
> same as if projecting from AFNI. However, when I try loading
> subject 1's dset onto subject 2's Icosahedron space, the
> alignment looks way off (not in the correct anatomical
> locations).
By dset you mean here an volumetric dataset? If that is the case then that is no surprise because the geometry of subject 2's surface has nothing to do with the volume from subject 1 and vice versa.

> Not sure if I'm doing something incorrect in
> 3dVol2Surf (although data looks fine when loading within
> subject data), or perhaps I'm missing a step.

It is a bit counterintuitive to understand but very easy to use once you understand the concept.

> On a seperate note I'm wondering if there is any way to convert
> a palette into a cmap file? I have a particular color map I
> want used in SUMA, however it is 16 colors, not continuous.
>
> I'd prefer a cmap file, but I also tried getting a .col file:
> Scale2Map -input file.1D.dset -cmapbd multi_rh.pal -cmap
> multiRH_n16 -anr 3.14 -nointerp > colormap.col
>
> I keep receiving an error: Could not find colormap multiRH_n16.
>
> My .pal file has this: ***PALETTES multiRH [16]
> The scale is from 1 to -1
>
> Any suggestions?
Hmm, could you upload your .pal file so I can take a look at this?

cheers,

z