@ skelband
"Why not just edit the orginal images before preparing the PDF or was that a deliberate troll?"
The explanation was in the comment you were responding to.
"***********************" isn't going to be "Bee", and "***" isn't going to be "Cholmondeley-Warner". Take detailed information about the font into account, and you can narrow down the possibilities considerably. Information is still leaked.
The point was that it can be a lot harder to securely redact documents, including images, than you might think. Editing the original images, as you suggest, might not be as secure as you might think.
As a quick experiment, redact "Bee", and then see if the same redaction would be enough to cover "Bay". Depending on your redaction, you might be surprised!
Paris, because of something to do with paying careful attention in relation to loss of privacy.