Thank you for your help,
in my case I want to estimate the directions via bedpostX and compare the results. For that I guess the different dwi's have to be in the same space. I thought doing eddy on all shells at the same time would solve this problem.
Sorry to ask again but I am not really into all this.
So first I have to coregister (flirt) the separate preprocessed dwi's and then apply the bvecs transform. How can I do this? Is there a way of doing it with fsl?
Maybe somebody has the same problem for comparing dwi's?
Thank you for any kind of help!