Re-run comparison to PSI + some minor fixes

[jokasimr]

I was looking at resolving the discrepancy in the result compared to PSI and as part of that I re-ran the Jochens Amor code to re-create the reference files that I was comparing against.

Surprisingly the results are now much closer. There are two possible reasons:

• I made a slight change in the command line arguments used to create the files, I don't think it should make a difference, but it might.
• The Amor reduction workflow has changed on their end and the changes made the results more similar.

In any case, I uploaded the new reference files to our file-server.

[jokasimr]

Comparison between the old and the new results:

Old results

New results

It's now only file 611 that has a real discrepancy. Still unclear why. Hopefully this insight will make the issue easier to debug. (Changing the sample rotation for 611 to 4.95 makes the curves overlap, might give a clue to understand what's going on.)

[jokasimr]

This is the commit that seems to have been the cause of the change: https://github.com/jochenstahn/amor/commit/140e3192ddb7e7f28acee87e2acaee65ce1332aa