Did it break apart the text elements from the background? Did it leave white blanked out areas that don’t match the sharpness of the black text that was lifted up? PDF scans might break up the scan (for compression purposes) but I don’t see how it would be smart enough to lift text and clean it up!