Hi, Jesper and other FSL experts

I have encountered a problem when running Eddy (FSL build 504) and would appreciate if you would help diagnoze the problem.

I have a set of diffusion data with 4 reptitions of blipUp and blipDn acquisitions. For each repetition, there are 128 directions + 11 b0s. So my data is arranged like this: 22b0s (b+) + 256dwi(b+) + 22b0s (b-) + 256dwi(b-). Before running Eddy, I ran Topup using 22b0s(b+) + 22b0s(b-), which was sucessful.

The index.txt file is 278 "1"s and 278 "2"s ( Am I right that I can use any two different numbers?)

Then, I fed all these  556 volumes to Eddy, togehter with the results from Topup and after aprproximately 24 hours, I got the results. The results are as follows: For the first 278 volumes, it seems both the eddy currents and susceptibility-induced distortion have been corrected and they look fine (the correction of eddy current is a little difficult to spot, as the eddy-related distortion is not huge in my data). But from 279th volume, the susceptibility-related distortion is obviously uncorrected at all and the images look like they have been stretched even more with large signal dropouts that do not show in the the uncorrected original images. So my question is:

(1) Is it possible that Eddy already merged the blipUp and blipDn in the first 278 volumes?
(2) If this results is mostly due to error, do you have any suggestions on how to fix it?
(3) I checked the previous posts and saw an option called "--very_verbose" and I could not find it anymore. Does that mean i am using a newer version? I have run Eddy twice and got the same problem and currently I am running it third time with "-v" option to see if it finishes sucessfully or aborts in the middle

Your help will be greatly appreciated!

Longchuan