I have a particular file generated on the mainframe (Z/OS). I found that the file is misinterpreted on recent notepad++ version (v220.127.116.11 in my case), it shows wrong characters and the number of characters is more than it should be. On an old version of notepad++ (v6.5.2 for instance), it works properly.
The following screenshot shows the difference between recent and old version : Capture
I made some tests and I finally found that the bug appears on notepad++ v6.5.5
To test it, use this file : File
Thanks for helping
N++ tries to autodetect the character encoding in version 6.5.5 and later. It can be turned off in Settings -> Preferences -> Misc.