When using some characters in annotation text, the characters are properly saved as part of the HTML portion of the annotation but not the text portion of the annotation even though the selected character set should include the characters. Characters in the plain text portion of the annotation text are instead replaced with unicode representations of the character.
Example
When creating a new annotation in Content Filtering > Annotations using non-ASCII characters
After saving the annotation the plain text representation has been changed to unicode character representations:
Release : 10.7.5
Component : Control Center
This is a defect in SMG 10.7.5 and the unicode representations of non-ASCII characters are being saved to the SMG configuration in place of the expected characters.
This issue has been addressed in the SMG 10.8.1 release.