Maybe this is intended - otherwise it should be considered to be a bug.

From some UML tool we generate reports in Word format with the graphics included full size but reduced to fit the page view and print.

After Word import the images in Confluence are reduced in size which makes them pretty useless, as object names on the images are not readable anymore.

Behavior is the same regardless of Word format (doc, docx).

Is that an intended behavior and is there a workaround? I try to avoid to replace every imported image manually by its full size image.

Version is Confluence 6.1.0