[MINC-users] minctracc angles problem

[Simon Fristed Eskildsen]

Andrew Janke wrote:
> try -identity... ;)
Well...identity is not an option in this case :-/

> Either that or there is some experimental code in there that uses
> quaternions in stead of Euler Angles for the optimisation, this is not
> limited to rotations of pi/2.  To use it I would suggest the following
> 
>    minctracc -quaternions -identity
The quaternions do not seem to do a better job (if any change at all).

I ended up providing a rough transformation as initial guess, which put 
the source within the pi/2 range of target.

Best,
Simon