Dear experts,
I encountered a pitfall with displaying control points in freeView. When the T1 and T2 geometry differs (which can occur for example in hires stream where data are not conformed to 1mm3), it depends on which volume is active when control.dat is loaded. For example, when control.dat is loaded with active T2 (which has different geometry than T1/brainmask), the control points are displayed on the incorrect position of the volume. This affects also loading data on commandline. For example: freeview -v T1.mgz -v T2.mgz -c control.dat displays control points incorrectly. freeview -v T2.mgz -v T1.mgz -c control.dat displays control points correctly.
Note, also, that in case of differing geometry of loaded images, the user has to pay special attention to select proper reference volume (brainmask.mgz or other derived images) when control points are created. The choice of reference volume is noted in the official documentation but maybe it can be easily overlooked with serious consequences.
To summarize, 4 different cases can occur:
control.dat created with reference volume brainmask.mgz, loaded into freeView with active volume brainmask.mgz -> all OK - control points applied correctly and also displayed correctly control.dat created with reference volume brainmask.mgz, loaded into freeView with active volume T2.mgz -> control points applied correctly, but displayed incorrectly (so issue only with display) control.dat created with reference volume T2.mgz loaded into freeView with active volume T2.mgz -> control points applied incorrectly, but displayed correctly in freeView (serious case !!) control.dat created with reference volume T2.mgz displayed with active volume brainmask.mgz -> control points applied incorrectly and also displayed incorrectly
Antonin Skoch