Indeed you need a recent version of EDDY to get the rotated bvecs.
As for your issue, it seems to me that rather than it being a difference between servers A and B, could it not be a difference in the data? You say you are doing denoising, but have you checked that the program you are using is not somehow changing e.g. the header info? Maybe you could try running the same command on the same data in both servers to check that it is not a software issue?
> On 26 Aug 2017, at 18:32, Emily <[log in to unmask]> wrote:
> Hi all,
> I am trying to run topup/eddy on a single shell dMRI sequence with dual phase encoding (AP and PA). I'm also working with a multi-shell sequence but let's start with the basics. I first ran the data on server A and everything looked great - bvecs rotated, tractography is beautiful. But I decided to do denoising first so I transferred the raw niftis to server B and now running eddy after topup I am not getting rotated bvecs. I get the *.eddy_parameters file and unwarped volumes, but no bvecs.
> I've tried several versions of fsl, and fsl 5.0.10 does give rotated bvecs, but they aren't right - V1 file looks flipped in the X direction, but even when I try to manually correct that, the tractography is a mess. I've read about issues with radiological/neurological convention, but I must be having a different issue.
> I'm nearly positive that server A has fsl 5.0 installed, but I've tried several different ones on server B and having the issue of either no rotated or incorrectly rotated bvecs output. Sorry if I've missed a thread that answers this, but any advice you can offer is appreciated.