Questions

One whose text is real text. The quick test is to open it in any reader and try to select a word — if the selection highlights the word, this can edit it. If nothing highlights, the page is an image and there is no text there to change.

No — and it is verified, not just promised. After the edit, every character outside the part you changed is compared against its position in the original, down to a fraction of a point, and the two pages are also compared as images. If anything moved, no file is produced.

Yes — the font already embedded in your document, addressed the way the document addresses it. No typeface is installed, downloaded or substituted.

Nearly always because the font in that spot does not carry a character you need. PDFs usually embed only the characters they use, so a heading font may hold forty characters. You are shown which characters are missing and which that font does have.

No. A scan is a picture of a page. Reading text out of it would mean character recognition and rebuilding the document, which gives you a new document rather than your own with one word fixed.

No. Open it in a reader with the password, save an unlocked copy, and bring that.

Yes, and you are told what it means. A signature exists so a reader can prove the file has not been altered since it was signed; an edited copy no longer matches it. Keep the original if you may need to show it is untouched.

No. Adding twenty words where five were is a new layout, not a correction, and reflowing a page is the thing that makes edited PDFs look edited. Small, exact changes are what this is for.

It is held while you work on it and deleted half an hour later. Your upload is never written over — every result is a new document — and nothing is kept afterwards.

40 MB per document.

No. It runs in the browser, on any operating system, and there is nothing to download.

No.