The infinity value and changed shape of proc diffusion MRI

Dear developers,

Thanks for the effort to make TORTOISE effective.

I am using TORTOISE v3.20. I have AP and PA phase encoding diffusion MRI. But in our imaging protocal the volumes of two diffusion MRI is different. I have run DIFFPREP and then I run DR_BUDDI. The log shows that those two process finish successfully. But I notice that the final processed image (up_data and down_data) have different image shape comparied with original input one. And also, there are infinity value existing in processed diffusion MRI image, which leading to failure of fitting tensor (EstimateTensorNLLS). In order to deal with that, I extract one b0 volume and make sure that there is no infinity values and then perform bet command to get mask so that we can exclude infinity value.  My question would be that is that normal or something wrong happened ? 

Thanks,
Qi